RFC 9580
OpenPGP
OpenPGP 🎀
Okay bestie, let's start with the ID card for RFC 9580 💖
- Title: OpenPGP
- RFC number: 9580
- Authors: P. Wouters, Ed. (Aiven), D. Huigens (Proton AG), J. Winter (Sequoia PGP), and Y. Niibe (FSIJ). The Ed. means P. Wouters is the editor 💅
- Category: Standards Track
- Date: July 2024
- ISSN: 2070-1721
- Obsoletes: RFC 4880 ("OpenPGP Message Format"), RFC 5581 ("The Camellia Cipher in OpenPGP"), and RFC 6637 ("Elliptic Curve Cryptography (ECC) in OpenPGP") 😭
Abstract ✨
Okay so this doc is all about the message formats used in OpenPGP. OpenPGP gives you encryption using public key or symmetric cryptographic algorithms, plus digital signatures, compression, and key management. It's giving full toolkit 🧰
The goal here is to publish everything you need to build interoperable apps based on the OpenPGP format. It is not a step-by-step cookbook for writing an app, though! It only covers the format and the methods you need to read, check, generate, and write conforming packets that cross any network. Storage and implementation questions are out of scope 🙈 It does, however, talk about the implementation issues you need to know about to avoid security flaws. Very important, no cap 🔥
This document obsoletes RFC 4880, RFC 5581, and RFC 6637 🦋
Status of This Memo 💜
This is an Internet Standards Track document 🌈
It's a product of the Internet Engineering Task Force (IETF), so it represents the consensus of the IETF community. It went through public review and got approved for publication by the Internet Engineering Steering Group (IESG). If you want more info on Internet Standards, check Section 2 of RFC 7841 🍓
Info about the current status of this document, any errata, and how to give feedback on it lives at https://www.rfc-editor.org/info/rfc9580 ⭐
Copyright Notice 🌸
Copyright (c) 2024 IETF Trust and the persons identified as the document authors. All rights reserved. 💅
This document is subject to BCP 78 and the IETF Trust's Legal Provisions Relating to IETF Documents (https://trustee.ietf.org/license-info) as they stand on the publication date. Please actually read these, bestie, because they explain your rights and restrictions for this document 🫶 Code Components pulled out of this document must include the Revised BSD License text described in Section 4.e of the Trust Legal Provisions, and they come without warranty, as the Revised BSD License describes.
This document may contain material from IETF Documents or IETF Contributions published or made publicly available before November 10, 2008. The people who control the copyright in some of that material may not have given the IETF Trust permission to allow modifications outside the IETF Standards Process. So unless you get an adequate license from those copyright holders, this document may not be modified outside the IETF Standards Process, and derivative works of it may not be created outside that process. The exceptions are formatting it for publication as an RFC or translating it into languages other than English 🎀
Table of Contents 🧚 picks up in the next part, so stay tuned!
Okay bestie, this is the whole map of the RFC 📑✨. Every section is in order with its number exactly as written, so you can find your way around 🗺️💅
1. Introduction
1.1. Terms
2. General Functions
2.1. Confidentiality via Encryption
2.2. Authentication via Digital Signature
2.3. Compression
2.4. Conversion to Base64
2.5. Signature-Only Applications
3. Data Element Formats
3.1. Scalar Numbers
3.2. Multiprecision Integers
3.2.1. Using MPIs to Encode Other Data
3.3. Key IDs and Fingerprints
3.4. Text
3.5. Time Fields
3.6. Keyrings
3.7. String-to-Key (S2K) Specifier
3.7.1. S2K Specifier Types
3.7.1.1. Simple S2K
3.7.1.2. Salted S2K
3.7.1.3. Iterated and Salted S2K
3.7.1.4. Argon2
3.7.2. S2K Usage
3.7.2.1. Secret Key Encryption
3.7.2.2. Symmetric Key Message Encryption
4. Packet Syntax
4.1. Overview
4.2. Packet Headers
4.2.1. OpenPGP Format Packet Lengths
4.2.1.1. 1-Octet Lengths
4.2.1.2. 2-Octet Lengths
4.2.1.3. 5-Octet Lengths
4.2.1.4. Partial Body Lengths
4.2.2. Legacy Format Packet Lengths
4.2.3. Packet Length Examples
4.3. Packet Criticality
5. Packet Types
5.1. Public Key Encrypted Session Key Packet (Type ID 1)
5.1.1. Version 3 Public Key Encrypted Session Key Packet
Format
5.1.2. Version 6 Public Key Encrypted Session Key Packet
Format
5.1.3. Algorithm-Specific Fields for RSA Encryption
5.1.4. Algorithm-Specific Fields for Elgamal Encryption
5.1.5. Algorithm-Specific Fields for ECDH Encryption
5.1.6. Algorithm-Specific Fields for X25519 Encryption
5.1.7. Algorithm-Specific Fields for X448 Encryption
5.1.8. Notes on PKESK
5.2. Signature Packet (Type ID 2)
5.2.1. Signature Types
5.2.1.1. Binary Signature (Type ID 0x00) of a Document
5.2.1.2. Text Signature (Type ID 0x01) of a Canonical
Document
5.2.1.3. Standalone Signature (Type ID 0x02)
5.2.1.4. Generic Certification Signature (Type ID 0x10) of a
User ID and Public Key Packet
5.2.1.5. Persona Certification Signature (Type ID 0x11) of a
User ID and Public Key Packet
5.2.1.6. Casual Certification Signature (Type ID 0x12) of a
User ID and Public Key Packet
5.2.1.7. Positive Certification Signature (Type ID 0x13) of
a User ID and Public Key Packet
5.2.1.8. Subkey Binding Signature (Type ID 0x18)
5.2.1.9. Primary Key Binding Signature (Type ID 0x19)
5.2.1.10. Direct Key Signature (Type ID 0x1F)
5.2.1.11. Key Revocation Signature (Type ID 0x20)
5.2.1.12. Subkey Revocation Signature (Type ID 0x28)
5.2.1.13. Certification Revocation Signature (Type ID 0x30)
5.2.1.14. Timestamp Signature (Type ID 0x40)
5.2.1.15. Third-Party Confirmation Signature (Type ID 0x50)
5.2.1.16. Reserved (Type ID 0xFF)
5.2.2. Version 3 Signature Packet Format
5.2.3. Versions 4 and 6 Signature Packet Formats
5.2.3.1. Algorithm-Specific Fields for RSA Signatures
5.2.3.2. Algorithm-Specific Fields for DSA or ECDSA
Signatures
5.2.3.3. Algorithm-Specific Fields for EdDSALegacy
Signatures (Deprecated)
5.2.3.4. Algorithm-Specific Fields for Ed25519 Signatures
5.2.3.5. Algorithm-Specific Fields for Ed448 Signatures
5.2.3.6. Notes on Signatures
5.2.3.7. Signature Subpacket Specification
5.2.3.8. Signature Subpacket Types
5.2.3.9. Notes on Subpackets
5.2.3.10. Notes on Self-Signatures
5.2.3.11. Signature Creation Time
5.2.3.12. Issuer Key ID
5.2.3.13. Key Expiration Time
5.2.3.14. Preferred Symmetric Ciphers for v1 SEIPD
5.2.3.15. Preferred AEAD Ciphersuites
5.2.3.16. Preferred Hash Algorithms
5.2.3.17. Preferred Compression Algorithms
5.2.3.18. Signature Expiration Time
5.2.3.19. Exportable Certification
5.2.3.20. Revocable
5.2.3.21. Trust Signature
5.2.3.22. Regular Expression
5.2.3.23. Revocation Key (Deprecated)
5.2.3.24. Notation Data
5.2.3.25. Key Server Preferences
5.2.3.26. Preferred Key Server
5.2.3.27. Primary User ID
5.2.3.28. Policy URI
5.2.3.29. Key Flags
5.2.3.30. Signer's User ID
5.2.3.31. Reason for Revocation
5.2.3.32. Features
5.2.3.33. Signature Target
5.2.3.34. Embedded Signature
5.2.3.35. Issuer Fingerprint
5.2.3.36. Intended Recipient Fingerprint
5.2.4. Computing Signatures
5.2.4.1. Notes about Signature Computation
5.2.5. Malformed and Unknown Signatures
5.3. Symmetric Key Encrypted Session Key Packet (Type ID 3)
5.3.1. Version 4 Symmetric Key Encrypted Session Key Packet
Format
5.3.2. Version 6 Symmetric Key Encrypted Session Key Packet
Format
5.4. One-Pass Signature Packet (Type ID 4)
5.5. Key Material Packets
5.5.1. Key Packet Variants
5.5.1.1. Public Key Packet (Type ID 6)
5.5.1.2. Public Subkey Packet (Type ID 14)
5.5.1.3. Secret Key Packet (Type ID 5)
5.5.1.4. Secret Subkey Packet (Type ID 7)
5.5.2. Public Key Packet Formats
5.5.2.1. Version 3 Public Keys
5.5.2.2. Version 4 Public Keys
5.5.2.3. Version 6 Public Keys
5.5.3. Secret Key Packet Formats
5.5.4. Key IDs and Fingerprints
5.5.4.1. Version 3 Key ID and Fingerprint
5.5.4.2. Version 4 Key ID and Fingerprint
5.5.4.3. Version 6 Key ID and Fingerprint
5.5.5. Algorithm-Specific Parts of Keys
5.5.5.1. Algorithm-Specific Part for RSA Keys
5.5.5.2. Algorithm-Specific Part for DSA Keys
5.5.5.3. Algorithm-Specific Part for Elgamal Keys
5.5.5.4. Algorithm-Specific Part for ECDSA Keys
5.5.5.5. Algorithm-Specific Part for EdDSALegacy Keys
(Deprecated)
5.5.5.6. Algorithm-Specific Part for ECDH Keys
5.5.5.7. Algorithm-Specific Part for X25519 Keys
5.5.5.8. Algorithm-Specific Part for X448 Keys
5.5.5.9. Algorithm-Specific Part for Ed25519 Keys
5.5.5.10. Algorithm-Specific Part for Ed448 Keys
5.6. Compressed Data Packet (Type ID 8)
5.7. Symmetrically Encrypted Data Packet (Type ID 9)
5.8. Marker Packet (Type ID 10)
5.9. Literal Data Packet (Type ID 11)
5.9.1. Special Filename _CONSOLE (Deprecated)
5.10. Trust Packet (Type ID 12)
5.11. User ID Packet (Type ID 13)
5.12. User Attribute Packet (Type ID 17)
5.12.1. Image Attribute Subpacket
5.13. Symmetrically Encrypted and Integrity Protected Data Packet
(Type ID 18)
5.13.1. Version 1 Symmetrically Encrypted and Integrity
Protected Data Packet Format
5.13.2. Version 2 Symmetrically Encrypted and Integrity
Protected Data Packet Format
5.13.3. EAX Mode
5.13.4. OCB Mode
5.13.5. GCM Mode
5.14. Padding Packet (Type ID 21)
6. Base64 Conversions
6.1. Optional Checksum
6.1.1. An Implementation of the CRC24 in "C"
6.2. Forming ASCII Armor
6.2.1. Armor Header Line
6.2.2. Armor Headers
6.2.2.1. "Version" Armor Header
6.2.2.2. "Comment" Armor Header
6.2.2.3. "Hash" Armor Header
6.2.2.4. "Charset" Armor Header
6.2.3. Armor Tail Line
7. Cleartext Signature Framework
7.1. Cleartext Signed Message Structure
7.2. Dash-Escaped Text
7.3. Issues with the Cleartext Signature Framework
8. Regular Expressions
9. Constants
9.1. Public Key Algorithms
9.2. ECC Curves for OpenPGP
9.2.1. Curve-Specific Wire Formats
9.3. Symmetric Key Algorithms
9.4. Compression Algorithms
9.5. Hash Algorithms
9.6. AEAD Algorithms
10. Packet Sequence Composition
10.1. Transferable Public Keys
10.1.1. OpenPGP Version 6 Certificate Structure
10.1.2. OpenPGP Version 6 Revocation Certificate
10.1.3. OpenPGP Version 4 Certificate Structure
10.1.4. OpenPGP Version 3 Key Structure
10.1.5. Common Requirements
10.2. Transferable Secret Keys
10.3. OpenPGP Messages
10.3.1. Unwrapping Encrypted and Compressed Messages
10.3.2. Additional Constraints on Packet Sequences
10.3.2.1. Packet Versions in Encrypted Messages
10.3.2.2. Packet Versions in Signatures
10.4. Detached Signatures
11. Elliptic Curve Cryptography
11.1. ECC Curves
11.2. EC Point Wire Formats
11.2.1. SEC1 EC Point Wire Format
11.2.2. Prefixed Native EC Point Wire Format
11.2.3. Notes on EC Point Wire Formats
11.3. EC Scalar Wire Formats
11.3.1. EC Octet String Wire Format
11.3.2. EC Prefixed Octet String Wire Format
11.4. Key Derivation Function
11.5. ECDH Algorithm
11.5.1. ECDH Parameters
12. Notes on Algorithms
12.1. PKCS#1 Encoding in OpenPGP
12.1.1. EME-PKCS1-v1_5-ENCODE
12.1.2. EME-PKCS1-v1_5-DECODE
12.1.3. EMSA-PKCS1-v1_5
12.2. Symmetric Algorithm Preferences
12.2.1. Plaintext
12.3. Other Algorithm Preferences
12.3.1. Compression Preferences
12.3.1.1. Uncompressed
12.3.2. Hash Algorithm Preferences
12.4. RSA
12.5. DSA
12.6. Elgamal
12.7. EdDSA
12.8. Reserved Algorithm IDs
12.9. CFB Mode
12.10. Private or Experimental Parameters
12.11. Meta Considerations for Expansion
13. Security Considerations
13.1. SHA-1 Collision Detection
13.2. Advantages of Salted Signatures
13.3. Elliptic Curve Side Channels
13.4. Risks of a Quick Check Oracle
13.5. Avoiding Leaks from PKCS#1 Errors
13.6. Fingerprint Usability
13.7. Avoiding Ciphertext Malleability
13.8. Secure Use of the v2 SEIPD Session-Key-Reuse Feature
13.9. Escrowed Revocation Signatures
13.10. Random Number Generation and Seeding
13.11. Traffic Analysis
13.12. Surreptitious Forwarding
13.13. Hashed vs. Unhashed Subpackets
13.14. Malicious Compressed Data
14. Implementation Considerations
14.1. Constrained Legacy Fingerprint Storage for Version 6 Keys
15. IANA Considerations
15.1. Renamed Protocol Group
15.2. Renamed and Updated Registries
15.3. Removed Registry
15.4. Added Registries
15.5. Registration Policies
15.5.1. Registries That Use RFC Required
15.6. Designated Experts
15.6.1. Key and Signature Versions
15.6.2. Encryption Versions
15.6.3. Algorithms
15.6.3.1. Elliptic Curve Algorithms
15.6.3.2. Symmetric Key Algorithms
15.6.3.3. Hash Algorithms
16. References
16.1. Normative References
16.2. Informative References
Appendix A. Test Vectors
A.1. Sample Version 4 Ed25519Legacy Key
A.2. Sample Version 4 Ed25519Legacy Signature
A.3. Sample Version 6 Certificate (Transferable Public Key)
A.3.1. Hashed Data Stream for Signature Verification
A.4. Sample Version 6 Secret Key (Transferable Secret Key)
A.5. Sample Locked Version 6 Secret Key (Transferable Secret
Key)
A.5.1. Intermediate Data for Locked Primary Key
A.5.2. Intermediate Data for Locked Subkey
A.6. Sample Cleartext Signed Message
A.7. Sample Inline-Signed Message
A.8. Sample X25519-AEAD-OCB Encryption and Decryption A.8.1. Sample Version 6 Public Key Encrypted Session Key
Packet
A.8.2. X25519 Encryption/Decryption of the Session Key
A.8.3. Sample v2 SEIPD Packet
A.8.4. Decryption of Data
A.8.5. Complete X25519-AEAD-OCB Encrypted Packet Sequence
A.9. Sample AEAD-EAX Encryption and Decryption
A.9.1. Sample Version 6 Symmetric Key Encrypted Session Key
Packet
A.9.2. Starting AEAD-EAX Decryption of the Session Key
A.9.3. Sample v2 SEIPD Packet
A.9.4. Decryption of Data
A.9.5. Complete AEAD-EAX Encrypted Packet Sequence
A.10. Sample AEAD-OCB Encryption and Decryption
A.10.1. Sample Version 6 Symmetric Key Encrypted Session Key
Packet
A.10.2. Starting AEAD-OCB Decryption of the Session Key
A.10.3. Sample v2 SEIPD Packet
A.10.4. Decryption of Data
A.10.5. Complete AEAD-OCB Encrypted Packet Sequence
A.11. Sample AEAD-GCM Encryption and Decryption
A.11.1. Sample Version 6 Symmetric Key Encrypted Session Key
Packet
A.11.2. Starting AEAD-GCM Decryption of the Session Key
A.11.3. Sample v2 SEIPD Packet
A.11.4. Decryption of Data
A.11.5. Complete AEAD-GCM Encrypted Packet Sequence
A.12. Sample Messages Encrypted Using Argon2
A.12.1. V4 SKESK Using Argon2 with AES-128
A.12.2. V4 SKESK Using Argon2 with AES-192
A.12.3. V4 SKESK Using Argon2 with AES-256
Appendix B. Upgrade Guidance (Adapting Implementations from RFCs
4880 and 6637)
B.1. Terminology Changes
Appendix C. Errata Addressed by This Document
Acknowledgements
Authors' Addresses1. Introduction ✨
Okay bestie, so this doc is all about the message-exchange packet formats that OpenPGP uses to do encryption, decryption, signing, and key management. 🔐 It's a revision of [RFC4880] ("OpenPGP Message Format"), which is a revision of [RFC2440], which itself replaces [RFC1991] ("PGP Message Exchange Formats"). Glow-up after glow-up 💅
This document obsoletes [RFC4880] (OpenPGP), [RFC5581] (Camellia in OpenPGP), and [RFC6637] (Elliptic Curves in OpenPGP). 🙈 At the time of writing, this document incorporates all outstanding verified errata, which are listed in Appendix C. Errata are the receipts, and we are keeping them all 🧾✨
If you've already implemented those older specs, you should totally check out Appendix B for pointers on what changed. It's giving "here's your cheat sheet" 📝🎀
1.1. Terms
The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in BCP 14 [RFC2119] [RFC8174] when, and only when, they appear in all capitals, as shown here. 📏 Only the ALL CAPS ones are the official rule words, so don't get it twisted 🙃
The key words "Private Use", "Specification Required", and "RFC Required" that appear in this document when used to describe namespace allocation are to be interpreted as described in [RFC8126]. 🗂️
Some terminology in this doc has been glowed up from previous versions of the OpenPGP spec. See Appendix B.1 for more details. ✨🦋
2. General Functions
Okay so OpenPGP gives you data confidentiality and integrity for messages and data files by using public key and/or symmetric encryption, plus digital signatures. 🔥 It also has formats for encoding and transferring encrypted and/or signed messages. And it does key and certificate encoding and transfer too, though key storage and management are beyond the scope of this document. Not its job, and it knows it 💅
2.1. Confidentiality via Encryption
OpenPGP combines symmetric key encryption and (optionally) public key encryption to provide confidentiality. 🔒 When using public keys, the object is first encrypted using a symmetric key encryption algorithm. Each symmetric key is used only once, for a single object. A new "session key" is generated as a random number for each object (sometimes called a "session"). Since it's only used once, the session key is bound to the message and sent along with it. To protect the key, it's encrypted with the receiver's public key. The sequence goes like this: 🫶
- The sender creates a message.
- The sending OpenPGP implementation generates a random session key for this message.
- The session key is encrypted using each recipient's public key. These "encrypted session keys" start the message.
- The sending OpenPGP implementation optionally compresses the message and then encrypts it using a message key derived from the session key. The encrypted message forms the remainder of the OpenPGP Message. 💌
- The receiving OpenPGP implementation decrypts the session key using the recipient's private key. 🗝️
- The receiving OpenPGP implementation decrypts the message using the message key derived from the session key. If the message was compressed, it will be decompressed. 🎁
When using symmetric key encryption, it's a similar process to the one above, but the session key is encrypted with a symmetric algorithm derived from a shared secret. 🤫
Both digital signature and confidentiality services may be applied to the same message. First, a signature is generated for the message and attached to it. Then the message plus signature gets encrypted using a symmetric message key derived from the session key. Finally, the session key is encrypted using public key encryption and prefixed to the encrypted block. The full glam 💖
2.2. Authentication via Digital Signature
The digital signature uses a cryptographic hash function and a public key algorithm that's capable of signing. ✍️ The sequence goes like this:
- The sender creates a message.
- The sending implementation generates a hash digest of the message.
- The sending implementation generates a signature from the hash digest using the sender's private key.
- The signature is attached to or transmitted alongside the message.
- The receiving implementation obtains a copy of the message and the message signature.
- The receiving implementation generates a new hash digest for the received message and verifies it using the message's signature. If the verification is successful, the message is accepted as authentic. ✅🌸
2.3. Compression
An OpenPGP implementation MAY support the compression of data. 🗜️ Many existing OpenPGP Messages are compressed. Implementers, like those working on constrained implementations that don't want to support compression, might want to consider at least implementing decompression. Just the unzipping, bestie 🧚
2.4. Conversion to Base64
OpenPGP's underlying representation for encrypted messages, signatures, keys, and certificates is a stream of arbitrary octets. 🌊 Some systems only allow blocks made of 7-bit, printable text. So for moving OpenPGP's raw binary octets through channels that aren't safe for raw binary data, there's a printable encoding defined. The raw 8-bit binary octet stream can be converted to a stream of printable ASCII characters using base64 encoding, in a format called "ASCII Armor" (see Section 6). It's the dress-up version of your data 🎀
Implementations SHOULD support base64 conversions. 🍓
2.5. Signature-Only Applications
OpenPGP is designed for applications that use both encryption and signatures, but there are a bunch of use cases that only need a signature-only implementation. 🖊️ Although this specification requires both encryption and signatures, it's reasonable for there to be subset implementations that are non-conformant only in that they leave out encryption support. Valid lil shortcut 🦋
3. Data Element Formats
This section describes the data elements used by OpenPGP. 🧩
3.1. Scalar Numbers
Scalar numbers are unsigned and always stored in big-endian format. 🔢 Using n[k] to refer to the kth octet being interpreted, the value of a 2-octet scalar is ((n[0] << 8) + n[1]). The value of a 4-octet scalar is ((n[0] << 24) + (n[1] << 16) + (n[2] << 8) + n[3]). Big-endian is giving "most significant byte goes first" energy 👑
3.2. Multiprecision Integers
Multiprecision Integers (MPIs) are unsigned integers used to hold large integers, like the ones used in cryptographic calculations. 🧮
An MPI consists of two pieces: a 2-octet scalar that is the length of the MPI in bits, followed by a string of octets that contain the actual integer. 📦
These octets form a big-endian number, and a big-endian number can be made into an MPI by prefixing it with the appropriate length. ✨
Examples:
(Note that all numbers in the octet strings identified by square brackets are in hexadecimal.) 🔟
The string of octets [00 00] forms an MPI with the value 0.
The string of octets [00 01 01] forms an MPI with the value 1.
The string [00 09 01 FF] forms an MPI with the value 511.
Additional rules:
- The size of an MPI is
((MPI.length + 7) / 8) + 2octets. - The length field of an MPI describes the length starting from its most significant non-zero bit. Thus, the MPI
[00 02 01]is not formed correctly. It should be[00 01 01]. When parsing an MPI in a version 6 Key, Signature, or Public Key Encrypted Session Key (PKESK) packet, the implementation MUST check that the encoded length matches the length starting from the most significant non-zero bit; if it doesn't match, reject the packet as malformed. 🚫 Length has to match the vibe, no exceptions 💀 - Unused bits of an MPI MUST be zero. 🫧
Quick reminder: the length field counts from the first non-zero bit, not from wherever you started. Count right or get rejected 🙅♀️
3.2.1. Using MPIs to Encode Other Data
Note that in some places, MPIs are used to encode non-integer data, like an elliptic curve (EC) point (see Section 11.2) or an octet string of known, fixed length (see Section 11.3). 🌈 The wire representation is the same: 2 octets of length in bits counted from the first non-zero bit, followed by the smallest series of octets that can represent the value while stripping off any leading zero octets. It's the same outfit, just a different occasion 🧥
3.3. Key IDs and Fingerprints
A Key ID is an 8-octet scalar that identifies a key. 🪪 Implementations SHOULD NOT assume that Key IDs are unique. A fingerprint is more likely to be unique than a Key ID. 💜 The fingerprint and Key ID of a key are calculated differently depending on the version of the key.
Section 5.5.4 describes how Key IDs and Fingerprints are formed. 📖
3.4. Text
Unless otherwise specified, the character set for text is the UTF-8 [RFC3629] encoding of Unicode [ISO10646]. 🌍💬
3.5. Time Fields
A time field is an unsigned 4-octet number containing the number of seconds elapsed since midnight, 1 January 1970 UTC. ⏰ Epoch vibes, all the way 🕛
3.6. Keyrings
A keyring is a collection of one or more keys in a file or database. Typically, a keyring is just a sequential list of keys, but it can be any suitable database. 🗃️ It's beyond the scope of this specification to discuss the details of keyrings or other databases. Not our tea ☕
3.7. String-to-Key (S2K) Specifier
A string-to-key (S2K) Specifier is used to convert a passphrase string into a symmetric key encryption/decryption key. 🔑 Passphrases that need S2K conversion are currently used in two places: to encrypt the secret part of private keys, and for symmetrically encrypted messages. Two spots, both very important, both very iconic 👑🔐
3.7.1. S2K Specifier Types
There are four types of S2K Specifiers currently specified, plus some reserved values 💅
+=========+==============+===============+==============+===========+
| ID | S2K Type | S2K Field | Generate? | Reference |
| | | Size | | |
| | | (Octets) | | |
+=========+==============+===============+==============+===========+
| 0 | Simple S2K | 2 | No | Section |
| | | | | 3.7.1.1 |
+---------+--------------+---------------+--------------+-----------+
| 1 | Salted S2K | 10 | Only when | Section |
| | | | string is | 3.7.1.2 |
| | | | high entropy | |
+---------+--------------+---------------+--------------+-----------+
| 2 | Reserved | - | No | |
| | value | | | |
+---------+--------------+---------------+--------------+-----------+
| 3 | Iterated and | 11 | Yes | Section |
| | Salted S2K | | | 3.7.1.3 |
+---------+--------------+---------------+--------------+-----------+
| 4 | Argon2 | 20 | Yes | Section |
| | | | | 3.7.1.4 |
+---------+--------------+---------------+--------------+-----------+
| 100-110 | Private or | - | As | |
| | Experimental | | appropriate | |
| | Use | | | |
+---------+--------------+---------------+--------------+-----------+
Table 1: OpenPGP String-to-Key (S2K) Types RegistryThe S2K Specifier Types get their own subsections below 🦋. If "Yes" is not in the "Generate?" column, that entry is only used for reading in backward-compatibility mode, and it is NOT RECOMMENDED... wait, let me be precise: it SHOULD NOT be used to generate new output 🙈
3.7.1.1. Simple S2K
Simple S2K directly hashes the string to produce the key data 🍓 The hashing is done like this:
Octet 0: 0x00
Octet 1: hash algorithmSimple S2K hashes the passphrase to make the session key ✨ How that works depends on the size of the session key (which depends on the cipher the session key is for) and the size of the hash algorithm's output. If the hash size is bigger than the session key size, the high-order (leftmost) octets of the hash get used as the key.
If the hash size is smaller than the key size, multiple instances of the hash context get created, enough to make the required key data 🌸 These instances are preloaded with 0, 1, 2, ... octets of zeros. So the first instance has no preloading, the second gets preloaded with 1 octet of zero, the third gets 2 octets of zeros, and so on.
As the data gets hashed, it's given to each hash context independently. Since the contexts were initialized differently, each one produces a different hash output 💜 Once the passphrase is hashed, the output from the multiple hashes is concatenated, first hash leftmost, to make the key data, and any extra octets on the right get thrown away.
3.7.1.2. Salted S2K
Salted S2K includes a "salt" value in the S2K Specifier (just some arbitrary data) that gets hashed along with the passphrase string. This helps prevent dictionary attacks 🛡️
Octet 0: 0x01
Octet 1: hash algorithm
Octets 2-9: 8-octet salt valueSalted S2K is exactly like Simple S2K, except the input to the hash function(s) is the 8 octets of salt from the S2K Specifier, followed by the passphrase 🎀
3.7.1.3. Iterated and Salted S2K
Iterated and Salted S2K includes both a salt and an octet count 🧚 The salt gets combined with the passphrase, and the resulting value is repeated and then hashed. This makes the work an attacker has to do to try dictionary attacks even bigger 💀
Octet 0: 0x03
Octet 1: hash algorithm
Octets 2-9: 8-octet salt value
Octet 10: count; a 1-octet coded valueThe count is coded into a 1-octet number using this formula:
#define EXPBIAS 6
count = ((Int32)16 + (c & 15)) << ((c >> 4) + EXPBIAS);The formula above is described in [C99], where "Int32" is a type for a 32-bit integer, and the variable "c" is the coded count, which is octet 10 🫶
Iterated and Salted S2K hashes the passphrase and salt data multiple times. The total number of octets to be hashed is set by the encoded count in the S2K Specifier. Heads up: the resulting count value is an octet count of how many octets get hashed, not an iteration count 👑
Initially, one or more hash contexts are set up the same way as the other S2K algorithms, depending on how many octets of key data are needed. Then the salt, followed by the passphrase data, is repeatedly processed as input to each hash context until the number of octets in the octet count has been hashed. The input gets truncated to the octet count, except if the octet count is less than the initial size of the salt plus passphrase. So at least one full copy of the salt plus passphrase is given to each hash context no matter what the octet count is 😭 After the hashing is done, the key data is produced from the hash digest(s), the same way it's produced for the other S2K algorithms.
3.7.1.4. Argon2
This S2K method hashes the passphrase using Argon2, as specified in [RFC9106] 🌈 It provides memory hardness, which further protects the passphrase against brute-force attacks. Literally iconic 🔥
Octet 0: 0x04
Octets 1-16: 16-octet salt value
Octet 17: 1-octet number of passes t
Octet 18: 1-octet degree of parallelism p
Octet 19: 1-octet encoded_m, specifying the exponent of
the memory sizeThe salt SHOULD be unique for each passphrase 🦋
The number of passes t and the degree of parallelism p MUST be non-zero.
The memory size m is 2^(encoded_m) kibibytes (KiB) of RAM. The encoded memory size MUST be a value from 3+ceil(log_2(p)) to 31, such that the decoded memory size m is a value from 8*p to 2^31. Note that memory-hardness size is indicated in KiB, not octets 🧠
Argon2 is invoked with the passphrase as P, the salt as S, the values of t, p, and m as described above, the required key size as the tag length T, 0x13 as the version v, and Argon2id as the type 💖
For the recommended values of t, p, and m, see Section 4 of [RFC9106]. If the recommended value of m for a given application is not a power of 2, it is RECOMMENDED to round up to the next power of 2 if the resulting performance would be acceptable. Otherwise, round down (keeping in mind that m must be at least 8*p) ✨
As an example, with the first recommended option (t=1, p=4, m=2^21), the full S2K Specifier would be:
04 XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX
XX 01 04 15where XX represents a random octet of salt 🌷
3.7.2. S2K Usage
Simple S2K and Salted S2K Specifiers can be brute-forced when used with a low-entropy string, like the ones users usually type in 😬 On top of that, using Simple S2K can lead to key and initialization vector (IV) reuse (see Section 5.3). So when generating an S2K Specifier, an implementation MUST NOT use Simple S2K. Furthermore, an implementation SHOULD NOT generate a Salted S2K unless it knows the input string is high entropy (for example, it generated the string itself using a known good source of randomness) 🔐
It is RECOMMENDED that implementations use Argon2. If Argon2 is not available, Iterated and Salted S2K MAY be used if care is taken to use a high octet count and a strong passphrase. However, this method does not provide memory hardness, unlike Argon2. So be careful, bestie 🫣
3.7.2.1. Secret Key Encryption ✨
okay so this is where we figure out whether a secret key is locked up with a passphrase, and how 🔐💅 The very first octet after the public key material in a Secret Key packet (Section 5.5.3) is the S2K usage octet, and it tells you if the secret key material is passphrase protected at all. Bestie, this one little byte is doing SO much 🫶
If the S2K usage octet is zero, the secret key data is unprotected. 🌸 If it's non-zero, it tells you how to use a passphrase to unlock the secret key. Main character energy 👑
Implementations from before [RFC2440] marked a protected key by putting a Symmetric Cipher Algorithm ID (see Section 9.3) in the S2K usage octet 🙈 In that case, MD5 was always used to turn the passphrase into a key for the cipher algorithm. Vintage, but make it make sense 💀
Later implementations use one of the special values 253 (AEAD), 254 (CFB), or 255 (MalleableCFB) in the S2K usage octet 🦋 The S2K usage octet is then followed right away by a set of fields that describe how to turn a passphrase into a symmetric key that can unlock the secret material, plus any other parameters that matter for the encryption type.
The wire format fields also depend on the version of the enclosing OpenPGP packet 🧚 The table below, indexed by the S2K usage octet, sums up the details from Section 5.5.3.
In the table below, check(x) means the "2-octet checksum", which is the sum of all octets in x mod 65536 🍓 The info and packetprefix parameters are explained in full detail in Section 5.5.3. Also, the "Generate?" column header got shortened to "Gen?" here, lowkey to save space 💖
Table 2: OpenPGP Secret Key Encryption (S2K Usage Octet) Registry
+=========+============+============+==========================+====+
|S2K Usage|Shorthand |Encryption |Encryption |Gen?|
|Octet | |Parameter | | |
| | |Fields | | |
+=========+============+============+==========================+====+
|0 |Unprotected |- |*v3 or v4 keys:* |Yes |
| | | |[cleartext secrets || | |
| | | |check(secrets)] | |
| | | |*v6 keys:* [cleartext | |
| | | |secrets] | |
+---------+------------+------------+--------------------------+----+
|Known |LegacyCFB |IV |CFB(MD5(passphrase), |No |
|symmetric| | |secrets || check(secrets))| |
|cipher | | | | |
|algo ID | | | | |
|(see | | | | |
|Section | | | | |
|9.3) | | | | |
+---------+------------+------------+--------------------------+----+
|253 |AEAD |params- |AEAD(HKDF(S2K(passphrase),|Yes |
| | |length |info), secrets, | |
| | |(*v6-only*),|packetprefix) | |
| | |cipher-algo,| | |
| | |AEAD-mode, | | |
| | |S2K- | | |
| | |specifier- | | |
| | |length | | |
| | |(*v6-only*),| | |
| | |S2K- | | |
| | |specifier, | | |
| | |nonce | | |
+---------+------------+------------+--------------------------+----+
|254 |CFB |params- |CFB(S2K(passphrase), |Yes |
| | |length |secrets || SHA1(secrets)) | |
| | |(*v6-only*),| | |
| | |cipher-algo,| | |
| | |S2K- | | |
| | |specifier- | | |
| | |length | | |
| | |(*v6-only*),| | |
| | |S2K- | | |
| | |specifier, | | |
| | |IV | | |
+---------+------------+------------+--------------------------+----+
|255 |MalleableCFB|cipher-algo,|CFB(S2K(passphrase), |No |
| | |S2K- |secrets || check(secrets))| |
| | |specifier, | | |
| | |IV | | |
+---------+------------+------------+--------------------------+----+When you're emitting a secret key (with or without passphrase protection), an implementation MUST only produce data from a row with "Generate?" marked as "Yes" 🫶 Every row marked "No" is only described for backward compatibility (for reading version 4 and earlier keys only), and those rows MUST NOT be used to generate new output. Version 6 secret keys using these formats MUST be rejected. 🚫
Heads up: compared to a version 4 secret key, the parameters of a passphrase-protected version 6 secret key are stored with an extra pair of length counts, each one octet wide 🎀
Argon2 is only used with Authenticated Encryption with Associated Data (AEAD) (S2K usage octet 253) 🔥 An implementation MUST NOT create, and MUST reject as malformed, any Secret Key packet where the S2K usage octet is not AEAD (253) and the S2K Specifier Type is Argon2. Not a vibe, bestie 💀
3.7.2.2. Symmetric Key Message Encryption 🦋
OpenPGP can make a Symmetric Key Encrypted Session Key (SKESK) packet at the front of a message ✨ This lets S2K Specifiers handle the passphrase conversion, or lets you make messages with a mix of SKESK packets and PKESK packets. Which means a message can be decrypted with either a passphrase or a public key pair. Iconic flexibility 💅
Implementations from before [RFC2440] always used the International Data Encryption Algorithm (IDEA) with Simple S2K conversion when encrypting a message with a symmetric algorithm; see Section 5.7 🌈 IDEA MUST NOT be generated but MAY be consumed for backward compatibility. Old school, no cap 🍓
4. Packet Syntax 📦
This section describes the packets used by OpenPGP 🧚
4.1. Overview 🌸
An OpenPGP Message is built from a bunch of records that are usually called packets 💖 A packet is a chunk of data with a Type ID that says what it means. An OpenPGP Message, keyring, certificate, detached signature, and so on are all made up of a number of packets. Some packets can contain other OpenPGP packets inside them (for example, a compressed data packet, when uncompressed, contains OpenPGP packets) 🎀
Each packet is a packet header followed by the packet body. The packet header has variable length ✨
When you're handling a stream of packets, the length information in each packet header is the canonical source of packet boundaries 👑 An implementation handling a packet stream that wants to find the next packet MUST look for it at the precise offset indicated in the previous packet header.
Some packets also contain internal length indicators (for example, a subfield within the packet) 🫶 If a subfield length indicator inside a packet implies including octets outside the range indicated in the packet header, a parser MUST abort without writing outside the indicated range, and MUST treat the packet as malformed and unusable. 💀
An implementation MUST NOT interpret octets outside the range indicated in the packet header as part of the contents of the packet 🙈
4.2. Packet Headers 🦋
The first octet of the packet tells you the format of the rest of the header, and it encodes the Packet Type ID, which says what type of packet it is (see Section 5) ✨ The rest of the packet header is the length of the packet.
There are two packet formats: 1) the (current) OpenPGP packet format specified by this document and its predecessors [RFC4880] and [RFC2440] and 2) the Legacy packet format, used by implementations that predate any IETF specification of OpenPGP 🧚
Quick note: the most significant bit is the leftmost bit, called "bit 7". A mask for this bit is 0x80 in hexadecimal 🍓
+---------------+
Encoded Packet Type ID: |7 6 5 4 3 2 1 0|
+---------------+
OpenPGP format:
Bit 7 -- always one
Bit 6 -- always one
Bits 5 to 0 -- Packet Type ID
Legacy format:
Bit 7 -- always one
Bit 6 -- always zero
Bits 5 to 2 -- Packet Type ID
Bits 1 to 0 -- length-typeBit 6 of the first octet of the packet header tells you whether the packet is encoded in the OpenPGP or Legacy packet format 👀 The Legacy packet format MAY be used when consuming packets, to help with interoperability and accessing archived data. The Legacy packet format SHOULD NOT be used to generate new data, unless the recipient is known to only support the Legacy packet format. That last case is extremely unlikely, since the Legacy packet format was obsoleted by [RFC2440] in 1998 💀
An implementation that consumes and redistributes pre-existing OpenPGP data (such as Transferable Public Keys) may run into packets framed with the Legacy packet format 🌸 Such an implementation MAY either redistribute these packets in their Legacy format, or transform them to the current OpenPGP packet format before redistributing 💖
Note that Legacy format headers only have 4 bits for the Packet Type ID, so they can only encode Packet Type IDs less than 16, whereas the OpenPGP format headers can encode IDs as great as 63 🎀
4.2.1. OpenPGP Format Packet Lengths 📏
OpenPGP format packets have four possible ways of encoding length:
- A 1-octet Body Length header encodes packet lengths of up to 191 octets.
- A 2-octet Body Length header encodes packet lengths of 192 to 8383 octets.
- A 5-octet Body Length header encodes packet lengths of up to 4,294,967,295 (0xFFFFFFFF) octets in length. (This actually encodes a 4-octet scalar number.)
- When the length of the packet body is not known in advance by the issuer, Partial Body Length headers encode a packet of indeterminate length, effectively making it a stream 🌊
4.2.1.1. 1-Octet Lengths 🧚
A 1-octet Body Length header encodes a length of 0 to 191 octets. You can spot this type because the 1-octet value is less than 192. The body length is equal to:
bodyLen = 1st_octet;4.2.1.2. 2-Octet Lengths 🦋
A 2-octet Body Length header encodes a length of 192 to 8383 octets. It's recognized because its first octet is in the range 192 to 223. The body length is equal to:
bodyLen = ((1st_octet - 192) << 8) + (2nd_octet) + 1924.2.1.3. 5-Octet Lengths 🌈
A 5-octet Body Length header consists of a single octet holding the value 255, followed by a 4-octet scalar. The body length is equal to:
bodyLen = (2nd_octet << 24) | (3rd_octet << 16) |
(4th_octet << 8) | 5th_octetThis basic set of 1-octet, 2-octet, and 5-octet lengths is also used internally in some packets 💜
4.2.1.4. Partial Body Lengths
okay so partial body lengths are a whole vibe 🦋 A Partial Body Length header is 1 octet long and it encodes the length of only part of the data packet ✨ The length is a power of 2, from 1 to 1,073,741,824 (2 to the 30th power) 🔥 You can spot one by its 1-octet value: it's greater than or equal to 224 and less than 255 💅 The Partial Body Length is equal to:
partialBodyLen = 1 << (1st_octet & 0x1F);Each Partial Body Length header is followed by a chunk of the packet body data, and the header tells you how long that chunk is 🌸 Then another length header (1-octet, 2-octet, 5-octet, or partial) comes after that chunk 🫶 The last length header in the packet MUST NOT be a Partial Body Length header 😭 Partial Body Length headers may only be used for the non-final parts of the packet 💀
Also, the last Body Length header can be a zero-length header 🙈
An implementation MAY use Partial Body Lengths for data packets, whether they are literal, compressed, or encrypted 🎀 The first partial length MUST be at least 512 octets long 💜 Partial Body Lengths MUST NOT be used for any other packet types 🚫✨
4.2.2. Legacy Format Packet Lengths
A zero in bit 6 of the first octet of the packet means it's a Legacy packet format 🌈 Bits 1 and 0 of the first octet of a Legacy packet are the "length-type" field 🍓 Here's what length-type means in Legacy format packets:
- 0: The packet has a 1-octet length, and the header is 2 octets long 🌸
- 1: The packet has a 2-octet length, and the header is 3 octets long 🦋
- 2: The packet has a 4-octet length, and the header is 5 octets long ✨
- 3: The packet is of indeterminate length 🧚 The header is 1 octet long, and the implementation must determine how long the packet is 🙈 If the packet is in a file, it means the packet extends until the end of the file 📂 The OpenPGP format headers have a mechanism for precisely encoding data of indeterminate length, so use that instead 💅 An implementation MUST NOT generate a Legacy format packet with indeterminate length 😤 An implementation MAY interpret an indeterminate length Legacy format packet in order to deal with historic data or data generated by a legacy system that predates support for [RFC2440] 🕰️
4.2.3. Packet Length Examples
These examples show ways that OpenPGP format packets might encode the packet body lengths 🎀
- A packet body with length 100 may have its length encoded in one octet:
0x64. This is followed by 100 octets of data 🍓 - A packet body with length 1723 may have its length encoded in two octets:
0xC5, 0xFB. This header is followed by the 1723 octets of data 🦋 - A packet body with length 100000 may have its length encoded in five octets:
0xFF, 0x00, 0x01, 0x86, 0xA0💖
It might also be encoded in the following octet stream:
0xEF, first 32768 octets of data;0xE1, next 2 octets of data;0xE0, next 1 octet of data;0xF0, next 65536 octets of data; and0xC5, 0xDD, last 1693 octets of data.
This is just one possible encoding, and many variations are possible on the size of the Partial Body Length headers, as long as a regular Body Length header encodes the last portion of the data 🧚
Please note that in all of these explanations, the total length of the packet is the length of the header(s) plus the length of the body 💜
4.3. Packet Criticality
The Packet Type ID space is partitioned into critical packets and non-critical packets 👑 If an implementation encounters a critical packet where the packet type is unknown in a packet sequence, it MUST reject the whole packet sequence (see Section 10) 🚨 On the other hand, an unknown non-critical packet MUST be ignored 🌸
Packets with Type IDs from 0 to 39 are critical 🔒 Packets with Type IDs from 40 to 63 are non-critical 🦋
5. Packet Types 💖
The defined packet types are as follows 📋
+=======+==========+=====================+===========+===========+
| ID | Critical | Packet Type | Shorthand | Reference |
| | | Description | | |
+=======+==========+=====================+===========+===========+
| 0 | Yes | Reserved - this | | |
| | | Packet Type ID MUST | | |
| | | NOT be used | | |
+-------+----------+---------------------+-----------+-----------+
| 1 | Yes | Public Key | PKESK | Section |
| | | Encrypted Session | | 5.1 |
| | | Key Packet | | |
+-------+----------+---------------------+-----------+-----------+
| 2 | Yes | Signature Packet | SIG | Section |
| | | | | 5.2 |
+-------+----------+---------------------+-----------+-----------+
| 3 | Yes | Symmetric Key | SKESK | Section |
| | | Encrypted Session | | 5.3 |
| | | Key Packet | | |
+-------+----------+---------------------+-----------+-----------+
| 4 | Yes | One-Pass Signature | OPS | Section |
| | | Packet | | 5.4 |
+-------+----------+---------------------+-----------+-----------+
| 5 | Yes | Secret Key Packet | SECKEY | Section |
| | | | | 5.5.1.3 |
+-------+----------+---------------------+-----------+-----------+
| 6 | Yes | Public Key Packet | PUBKEY | Section |
| | | | | 5.5.1.1 |
+-------+----------+---------------------+-----------+-----------+
| 7 | Yes | Secret Subkey | SECSUBKEY | Section |
| | | Packet | | 5.5.1.4 |
+-------+----------+---------------------+-----------+-----------+
| 8 | Yes | Compressed Data | COMP | Section |
| | | Packet | | 5.6 |
+-------+----------+---------------------+-----------+-----------+
| 9 | Yes | Symmetrically | SED | Section |
| | | Encrypted Data | | 5.7 |
| | | Packet | | |
+-------+----------+---------------------+-----------+-----------+
| 10 | Yes | Marker Packet | MARKER | Section |
| | | | | 5.8 |
+-------+----------+---------------------+-----------+-----------+
| 11 | Yes | Literal Data Packet | LIT | Section |
| | | | | 5.9 |
+-------+----------+---------------------+-----------+-----------+
| 12 | Yes | Trust Packet | TRUST | Section |
| | | | | 5.10 |
+-------+----------+---------------------+-----------+-----------+
| 13 | Yes | User ID Packet | UID | Section |
| | | | | 5.11 |
+-------+----------+---------------------+-----------+-----------+
| 14 | Yes | Public Subkey | PUBSUBKEY | Section |
| | | Packet | | 5.5.1.2 |
+-------+----------+---------------------+-----------+-----------+
| 17 | Yes | User Attribute | UAT | Section |
| | | Packet | | 5.12 |
+-------+----------+---------------------+-----------+-----------+
| 18 | Yes | Symmetrically | SEIPD | Section |
| | | Encrypted and | | 5.13 |
| | | Integrity Protected | | |
| | | Data Packet | | |
+-------+----------+---------------------+-----------+-----------+
| 19 | Yes | Reserved (formerly | | Section |
| | | Modification | | 5.13.1 |
| | | Detection Code | | |
| | | Packet) | | |
+-------+----------+---------------------+-----------+-----------+
| 20 | Yes | Reserved | | |
+-------+----------+---------------------+-----------+-----------+
| 21 | Yes | Padding Packet | PADDING | Section |
| | | | | 5.14 |
+-------+----------+---------------------+-----------+-----------+
| 22-39 | Yes | Unassigned Critical | | |
| | | Packets | | |
+-------+----------+---------------------+-----------+-----------+
| 40-59 | No | Unassigned Non- | | |
| | | Critical Packets | | |
+-------+----------+---------------------+-----------+-----------+
| 60-63 | No | Private or | | |
| | | Experimental Use | | |
+-------+----------+---------------------+-----------+-----------+
Table 3: OpenPGP Packet Types RegistryThe labels in the "Shorthand" column are used for compact reference elsewhere in this document 🔖 They may also be used by implementations that provide debugging or inspection affordances for streams of OpenPGP packets 🔍
5.1. Public Key Encrypted Session Key Packet (Type ID 1) 🔐
Zero or more PKESK packets and/or SKESK packets (Section 5.3) precede an encryption container, which holds an Encrypted Message 📦 That container is either a Symmetrically Encrypted and Integrity Protected Data (SEIPD) packet or, for historic data, a Symmetrically Encrypted Data (SED) packet 🕰️ The message is encrypted with the session key, and the session key is itself encrypted and stored in the Encrypted Session Key packet(s) 🗝️ The encryption container is preceded by one Public Key Encrypted Session Key packet for each OpenPGP Key the message is encrypted to 💌 The recipient finds a session key that's encrypted to their public key, decrypts the session key, and then uses it to decrypt the message 🎉
The body of this packet starts with a 1-octet number giving the version number of the packet type 🔢 The currently defined versions are 3 and 6 ✨ The remainder of the packet depends on the version 🧩
The versions differ in how they identify the recipient key and in what they encode 🪪 The version of the PKESK packet MUST align with the version of the SEIPD packet (see Section 10.3.2.1) 💯 Any new version of the PKESK packet SHOULD be registered in the registry established in Section 10.3.2.1 📝
5.1.1. Version 3 Public Key Encrypted Session Key Packet Format ✨
okay so a v3 PKESK packet shows up right before a v1 SEIPD packet (see Section 5.13.1) 🎀. In historic data you can sometimes spot one sitting in front of a deprecated SED packet (see Section 5.7) 💀. BUT big rule: a v3 PKESK packet MUST NOT precede a v2 SEIPD packet (see Section 10.3.2.1) 🙅♀️. do not mix these up, bestie 😭
The v3 PKESK packet is made of:
- A 1-octet version number with value 3. 🌸
- An 8-octet number that gives the Key ID of the public key the session key is encrypted to. If the session key is encrypted to a subkey, then the Key ID of that subkey goes here instead of the primary key's Key ID 🔑. The Key ID can also be all zeros, for an "anonymous recipient" (see Section 5.1.8) 🦋.
- A 1-octet number giving the public key algorithm used. 💅
- A series of values comprising the encrypted session key. This part is algorithm specific and we'll get into it below 🧚.
The public key encryption algorithm (covered in the next sections 💜) gets passed two values:
- The session key.
- The 1-octet algorithm identifier that specifies the symmetric key encryption algorithm used to encrypt the v1 SEIPD packet in the next section ✨.
5.1.2. Version 6 Public Key Encrypted Session Key Packet Format 🌈
A v6 PKESK packet precedes a v2 SEIPD packet (see Section 5.13.2) 🫶. A v6 PKESK packet MUST NOT precede a v1 SEIPD packet or a deprecated SED packet (see Section 10.3.2.1) 🙈.
The v6 PKESK packet consists of the following fields:
- A 1-octet version number with value 6. 🍓
- A 1-octet size of the following two fields. This size may be zero, if the key version number field and the fingerprint field are omitted for an "anonymous recipient" (see Section 5.1.8) 👑.
- A 1-octet key version number. 🎀
- The fingerprint of the public key or subkey the session key is encrypted to. Note that the length N of the fingerprint for a version 4 key is 20 octets; for a version 6 key, N is 32 ✨.
- A 1-octet number giving the public key algorithm used. 💖
- A series of values comprising the encrypted session key. This is algorithm specific and described below 🦋.
The session key is encrypted according to the public key algorithm used, as described below. No symmetric key encryption algorithm identifier is passed to the public key algorithm for a v6 PKESK packet, because it's already included in the v2 SEIPD packet 💅.
5.1.3. Algorithm-Specific Fields for RSA Encryption 🔥
- MPI of RSA-encrypted value
m^e mod n.
To produce the value "m" in the formula above, first concatenate the following values:
- The 1-octet algorithm identifier, if it was passed (in the case of a v3 PKESK packet). 🌸
- The session key.
- A 2-octet checksum of the session key, equal to the sum of the session key octets, modulo 65536 🧮.
Then the above values are encoded using the PKCS#1 block encoding EME-PKCS1-v1_5, as described in Step 2 in Section 7.2.1 of [RFC8017] (see also Section 12.1.1) 🧁. When decoding "m" during decryption, an implementation should follow Step 3 in Section 7.2.2 of [RFC8017] (see also Section 12.1.2) 🫧.
Note that when an implementation forms several PKESK packets with one session key, making a message that can be decrypted by several keys, the implementation MUST make a new PKCS#1 encoding for each key. This defends against attacks such as those discussed in [HASTAD] 🛡️✨.
5.1.4. Algorithm-Specific Fields for Elgamal Encryption 🌷
- MPI of Elgamal (Diffie-Hellman) value
g^k mod p. - MPI of Elgamal (Diffie-Hellman) value
m * y^k mod p.
To produce the value "m" in the formula above, first concatenate the following values:
- The 1-octet algorithm identifier, if it was passed (in the case of a v3 PKESK packet). 🍒
- The session key.
- A 2-octet checksum of the session key, equal to the sum of the session key octets, modulo 65536 🧮.
Then the above values are encoded using the PKCS#1 block encoding EME-PKCS1-v1_5, as described in Step 2 in Section 7.2.1 of [RFC8017] (see also Section 12.1.1). When decoding "m" during decryption, an implementation should follow Step 3 in Section 7.2.2 of [RFC8017] (see also Section 12.1.2) 🌙.
Note that when an implementation forms several PKESK packets with one session key, making a message that can be decrypted by several keys, the implementation MUST make a new PKCS#1 encoding for each key. This defends against attacks such as those discussed in [HASTAD] 💜.
An implementation MUST NOT generate ElGamal v6 PKESK packets. 🚫 no exceptions, babe.
5.1.5. Algorithm-Specific Fields for ECDH Encryption 🦋
- MPI of an EC point representing an ephemeral public key in the point format associated with the curve, as specified in Section 9.2 ✨.
- A 1-octet size, followed by a symmetric key encoded using the method described in Section 11.5 🎀.
5.1.6. Algorithm-Specific Fields for X25519 Encryption 🌈
- 32 octets representing an ephemeral X25519 public key.
- A 1-octet size of the following fields.
- The 1-octet algorithm identifier, if it was passed (in the case of a v3 PKESK packet).
- The encrypted session key.
See Section 6.1 of [RFC7748] for more details on how the ephemeral public key and the shared secret are computed 🧠. The HMAC-based Key Derivation Function (HKDF) [RFC5869] is then used with SHA256 [RFC6234] and an info parameter of OpenPGP X25519 and no salt. The input of HKDF is the concatenation of these three values:
- 32 octets of the ephemeral X25519 public key from this packet.
- 32 octets of the recipient public key material.
- 32 octets of the shared secret.
The key produced by HKDF is used to encrypt the session key with AES-128 key wrap, as defined in [RFC3394] 🔐.
Note that, unlike Elliptic Curve Diffie-Hellman (ECDH), no checksum or padding is appended to the session key before key wrapping. Finally, note that, unlike the other public key algorithms, in the case of a v3 PKESK packet the symmetric algorithm ID is not encrypted. Instead, it is prepended to the encrypted session key in plaintext. In this case, the symmetric algorithm used MUST be AES-128, AES-192, or AES-256 (algorithm IDs 7, 8, or 9, respectively) 🫶.
5.1.7. Algorithm-Specific Fields for X448 Encryption 💎
- 56 octets representing an ephemeral X448 public key.
- A 1-octet size of the following fields.
- The 1-octet algorithm identifier, if it was passed (in the case of a v3 PKESK packet).
- The encrypted session key.
See Section 6.2 of [RFC7748] for more details on how the ephemeral public key and the shared secret are computed 🧠. HKDF [RFC5869] is then used with SHA512 [RFC6234] and an info parameter of OpenPGP X448 and no salt. The input of HKDF is the concatenation of these three values:
- 56 octets of the ephemeral X448 public key from this packet.
- 56 octets of the recipient public key material.
- 56 octets of the shared secret.
The key produced by HKDF is used to encrypt the session key with AES-256 key wrap, as defined in [RFC3394] 🔐.
Note that, unlike ECDH, no checksum or padding is appended to the session key before key wrapping. Finally, note that, unlike the other public key algorithms, in the case of a v3 PKESK packet the symmetric algorithm ID is not encrypted. Instead, it is prepended to the encrypted session key in plaintext. In this case, the symmetric algorithm used MUST be AES-128, AES-192, or AES-256 (algorithm ID 7, 8, or 9) 🌟.
5.1.8. Notes on PKESK 🙈
An implementation MAY accept or use a Key ID of all zeros, or an omitted key fingerprint, to hide the intended decryption key. In this case, the receiving implementation would try all available private keys, checking for a valid decrypted session key. This format helps reduce traffic analysis of messages 🕵️♀️💅.
5.2. Signature Packet (Type ID 2) ✍️
A Signature packet describes a binding between some public key and some data. The most common signatures are a signature of a file or a block of text, and a signature that is a certification of a User ID. It's giving main character energy 👑.
Three versions of Signature packets are defined. Version 3 provides basic signature information, while versions 4 and 6 provide an expandable format with subpackets that can specify more information about the signature 🌸.
For historical reasons, versions 1, 2, and 5 of the Signature packet are unspecified. Any new Signature packet version should be registered in the registry established in Section 10.3.2.2 📋.
An implementation MUST generate a version 6 signature when signing with a version 6 key. An implementation MUST generate a version 4 signature when signing with a version 4 key. Implementations MUST NOT create version 3 signatures; they MAY accept version 3 signatures. See Section 10.3.2.2 for more details about packet version correspondence between keys and signatures 🦋✨.
5.2.1. Signature Types ✨
okay so a signature can mean a bunch of different things, and the Signature Type ID tells you which meaning you're looking at 💅. the vagueness is honestly a feature and not a bug 🦋. OpenPGP gives the final say on validity to the receiver, so one signer's casual little nod might be more rigorous than some other authority's big official stamp. see Section 5.2.4 for how to compute and verify each type 🫶
+======+====================================+==================+
| ID | Name | Reference |
+======+====================================+==================+
| 0x00 | Binary Signature | Section 5.2.1.1 |
+------+------------------------------------+------------------+
| 0x01 | Text Signature | Section 5.2.1.2 |
+------+------------------------------------+------------------+
| 0x02 | Standalone Signature | Section 5.2.1.3 |
+------+------------------------------------+------------------+
| 0x10 | Generic Certification Signature | Section 5.2.1.4 |
+------+------------------------------------+------------------+
| 0x11 | Persona Certification Signature | Section 5.2.1.5 |
+------+------------------------------------+------------------+
| 0x12 | Casual Certification Signature | Section 5.2.1.6 |
+------+------------------------------------+------------------+
| 0x13 | Positive Certification Signature | Section 5.2.1.7 |
+------+------------------------------------+------------------+
| 0x18 | Subkey Binding Signature | Section 5.2.1.8 |
+------+------------------------------------+------------------+
| 0x19 | Primary Key Binding Signature | Section 5.2.1.9 |
+------+------------------------------------+------------------+
| 0x1F | Direct Key Signature | Section 5.2.1.10 |
+------+------------------------------------+------------------+
| 0x20 | Key Revocation Signature | Section 5.2.1.11 |
+------+------------------------------------+------------------+
| 0x28 | Subkey Revocation Signature | Section 5.2.1.12 |
+------+------------------------------------+------------------+
| 0x30 | Certification Revocation Signature | Section 5.2.1.13 |
+------+------------------------------------+------------------+
| 0x40 | Timestamp Signature | Section 5.2.1.14 |
+------+------------------------------------+------------------+
| 0x50 | Third-Party Confirmation Signature | Section 5.2.1.15 |
+------+------------------------------------+------------------+
| 0xFF | Reserved | Section 5.2.1.16 |
+------+------------------------------------+------------------+
Table 4: OpenPGP Signature Types Registryeach signature type's meaning gets its own little subsection below 💖
5.2.1.1. Binary Signature (Type ID 0x00) of a Document
this one means the signer owns it, created it, or certifies that it has not been modified 🌸
5.2.1.2. Text Signature (Type ID 0x01) of a Canonical Document
same energy as the binary one: the signer owns it, created it, or certifies it hasn't been modified ✨ the difference is that the signature gets calculated over the text data with its line endings converted to <CR><LF> 🎀
5.2.1.3. Standalone Signature (Type ID 0x02)
this one is a signature of only its own subpacket contents 🧚 it's calculated identically to a signature over a zero-length binary document. Version 3 Standalone signatures MUST NOT be generated and MUST be ignored. 🙈
5.2.1.4. Generic Certification Signature (Type ID 0x10) of a User ID and Public Key Packet
the issuer of this certification doesn't make any particular assertion about how well the certifier has checked that the owner of the key is actually the person described by the User ID 🤷♀️
5.2.1.5. Persona Certification Signature (Type ID 0x11) of a User ID and Public Key Packet
the issuer of this one has not done any verification at all that the key's owner is the User ID specified 💀
5.2.1.6. Casual Certification Signature (Type ID 0x12) of a User ID and Public Key Packet
the issuer did some casual verification of the identity claim 👀 a little check, but not a big deal
5.2.1.7. Positive Certification Signature (Type ID 0x13) of a User ID and Public Key Packet
the issuer did substantial verification of the identity claim 👑 this is the big-boss level one
most OpenPGP implementations make their "key signatures" as generic (Type ID 0x10) certifications 📝 some implementations can issue 0x11-0x13 certifications, but few actually differentiate between the types 😭
5.2.1.8. Subkey Binding Signature (Type ID 0x18)
this is a statement by the top-level signing key saying it owns the subkey 🔑 it's calculated directly on the primary key and subkey, and not on any User ID or other packets. a signature that binds a signing subkey MUST have an Embedded Signature subpacket in this binding signature, and that subpacket has to contain a 0x19 signature made by the signing subkey on the primary key and subkey 🫶
5.2.1.9. Primary Key Binding Signature (Type ID 0x19)
this one is a statement by a signing subkey saying that it's owned by the primary key 💜 it's calculated the same way as a Subkey Binding signature (Type ID 0x18): directly on the primary key and subkey, and not on any User ID or other packets 🦋
5.2.1.10. Direct Key Signature (Type ID 0x1F)
this is calculated directly on a key 🗝️ it binds the information in the Signature subpackets to the key, and it's the right home for subpackets that describe the key itself, like the Key Flags subpacket or the (deprecated) Revocation Key subpacket. it's also the right move for statements that non-self certifiers want to make about the key itself, rather than about the binding between a key and a name 🌈
5.2.1.11. Key Revocation Signature (Type ID 0x20)
this one is calculated directly on the key being revoked 🚫 a revoked key is not to be used. only Revocation Signatures by the key being revoked, or by a (deprecated) Revocation Key, should be considered valid Revocation Signatures 🙅♀️
5.2.1.12. Subkey Revocation Signature (Type ID 0x28)
this is calculated directly on the primary key and the subkey being revoked 🥀 a revoked subkey is not to be used. only Revocation Signatures by the top-level signature key that's bound to this subkey, or by a (deprecated) Revocation Key, should be considered valid Revocation Signatures 💅
5.2.1.13. Certification Revocation Signature (Type ID 0x30)
this one revokes an earlier User ID certification signature (Type IDs 0x10 through 0x13) or Direct Key signature (Type ID 0x1F) 🪦 it should be issued by the same key that issued the revoked signature, or by a (deprecated) Revocation Key. the signature is computed over the same data as the certification it revokes, and it should have a later creation date than that certification ⏰
5.2.1.14. Timestamp Signature (Type ID 0x40)
this signature is only meaningful for the timestamp contained in it ⏱️ that's it, that's the whole vibe 🌸
5.2.1.15. Third-Party Confirmation Signature (Type ID 0x50)
this is a signature over another OpenPGP Signature packet 📜 it's basically analogous to a notary seal on the signed data, so very official and very iconic 💅 a Third-Party Confirmation signature SHOULD include a Signature Target subpacket that identifies the confirmed signature 🎀
5.2.1.16. Reserved (Type ID 0xFF)
An implementation MUST NOT create any signature with this type and MUST NOT validate any signature made with this type. 🚨 see Section 5.2.4.1 for more details 🙈
5.2.2. Version 3 Signature Packet Format
the body of a version 3 Signature packet has these bits in it 🍓
- A 1-octet version number with value 3.
- A 1-octet length of the following hashed material; it MUST be 5:
- A 1-octet Signature Type ID.
- A 4-octet creation time.
- An 8-octet Key ID of the signer.
- A 1-octet public key algorithm.
- A 1-octet hash algorithm.
- A 2-octet field holding left 16 bits of the signed hash value.
- One or more MPIs comprising the signature. This portion is algorithm specific, as described below.
okay so the concatenation of the data to be signed, the signature type, and the creation time from the Signature packet (5 additional octets) gets hashed 🧁 the resulting hash value is what goes into the signature algorithm. the high 16 bits (first two octets) of the hash are included in the Signature packet, which gives you a way to reject some invalid signatures without doing a full signature verification ⚡
algorithm-specific fields for RSA signatures:
- MPI of RSA signature value
m^d mod n.
algorithm-specific fields for DSA signatures:
- MPI of DSA value
r. - MPI of DSA value
s.
the signature calculation is based on a hash of the signed data, as described above 🌈 the details are different for DSA signatures than for RSA signatures, so check out Sections 5.2.3.1 and 5.2.3.2 🦋
5.2.3. Versions 4 and 6 Signature Packet Formats
the body of a version 4 or version 6 Signature packet contains 🎀
- A 1-octet version number. This is 4 for version 4 signatures and 6 for version 6 signatures.
- A 1-octet Signature Type ID.
- A 1-octet public key algorithm.
- A 1-octet hash algorithm.
- A scalar octet count for the hashed subpacket data that follows this field. For a version 4 signature, this is a 2-octet field. For a version 6 signature, this is a 4-octet field. note that this is the length in octets of all of the hashed subpackets, so an implementation's pointer incremented by this number will skip right over the hashed subpackets 🏃♀️
- A hashed subpacket data set (zero or more subpackets).
- A scalar octet count for the unhashed subpacket data that follows this field. For a version 4 signature, this is a 2-octet field. For a version 6 signature, this is a 4-octet field. note that this is the length in octets of all of the unhashed subpackets, so an implementation's pointer incremented by this number will skip over the unhashed subpackets 🌸
- An unhashed subpacket data set (zero or more subpackets).
- A 2-octet field holding the left 16 bits of the signed hash value.
- Only for version 6 signatures, a variable-length field containing:
- A 1-octet salt size. The value MUST match the value defined for the hash algorithm as specified in Table 23.
- The salt, which is a random value of the specified size. 🧂
- One or more MPIs comprising the signature. This portion is algorithm specific.
5.2.3.1. Algorithm-Specific Fields for RSA Signatures
- MPI of RSA signature value
m^d mod n.
with RSA signatures, the hash value is encoded using PKCS#1 encoding type EMSA-PKCS1-v1_5, as described in Section 9.2 of [RFC8017] (see also Section 12.1.3) 🔐 this requires inserting the hash value as an octet string into an ASN.1 structure. the object identifier (OID) for the hash algorithm itself is also included in the structure, so check the OIDs in Table 24 💖
5.2.3.2. Algorithm-Specific Fields for DSA or ECDSA Signatures ✨
okay so for DSA and ECDSA signatures, the packet carries two MPIs:
- MPI of DSA or ECDSA value r. 🎀
- MPI of DSA or ECDSA value s. 🎀
ngl this part is strict. A version 3 signature MUST NOT be created and MUST NOT be used with the Elliptic Curve Digital Signature Algorithm (ECDSA). 💀 That's a hard no, bestie.
A DSA signature MUST use a hash algorithm with a digest size of at least the number of bits of q, the group generated by the DSA key's generator value. 🔥
If the output size of the chosen hash is larger than the number of bits of q, the hash result gets trimmed to fit by keeping only the leftmost bits, equal to the number of bits of q. This (possibly truncated) hash function result is treated as a number and used directly in the DSA signature algorithm. 🦋
An ECDSA signature MUST use a hash algorithm with a digest size of at least the curve's fsize value (see Section 9.2), except in the case of NIST P-521, for which at least a 512-bit hash algorithm MUST be used. 👑
5.2.3.3. Algorithm-Specific Fields for EdDSALegacy Signatures (Deprecated) 🥀
this one is deprecated, so it's giving "history lesson" energy 📜
- Two MPI-encoded values, whose contents and formatting depend on the choice of curve used (see Section 9.2.1). 🌸
A version 3 signature MUST NOT be created and MUST NOT be used with EdDSALegacy. 🙈 no version 3 for this one either.
An EdDSALegacy signature MUST use a hash algorithm with a digest size of at least the curve's fsize value (see Section 9.2). A verifying implementation MUST reject any EdDSALegacy signature that uses a hash algorithm with a smaller digest size. 💅 the verifier is the bouncer here, and it's not letting anyone with a tiny hash in.
5.2.3.3.1. Algorithm-Specific Fields for Ed25519Legacy Signatures (Deprecated) 🧚
the two MPIs for Ed25519Legacy are the octet strings R and S from the Edwards-curve Digital Signature Algorithm (EdDSA), described in [RFC8032]. basically the same math, just wrapped up in MPI packaging. 🎁
- MPI of an EC point R, represented as a (non-prefixed) native (little-endian) octet string up to 32 octets. 🌈
- MPI of EdDSA value S, also in (non-prefixed) native (little-endian) format with a length up to 32 octets. 🌈
Ed25519Legacy MUST NOT be used in Signature packets version 6 or above. 🚫 Versions 6 and up are a no-go for this one, periodt.
5.2.3.4. Algorithm-Specific Fields for Ed25519 Signatures ✨
- 64 octets of the native signature. 🍓
for more details, check out Section 12.7. 🔍
A version 3 signature MUST NOT be created and MUST NOT be used with Ed25519. 🙅♀️ version 3 is not invited to this party.
An Ed25519 signature MUST use a hash algorithm with a digest size of at least 256 bits. A verifying implementation MUST reject any Ed25519 signature that uses a hash algorithm with a smaller digest size. 🛡️ the verifier is so protective of its hashes, it's iconic.
5.2.3.5. Algorithm-Specific Fields for Ed448 Signatures ✨
- 114 octets of the native signature. 💜
for more details, see Section 12.7 again, because it's a big deal. 📖
A version 3 signature MUST NOT be created and MUST NOT be used with Ed448. 🙈 still no version 3, sorry not sorry.
An Ed448 signature MUST use a hash algorithm with a digest size of at least 512 bits. A verifying implementation MUST reject any Ed448 signature that uses a hash algorithm with a smaller digest size. 🔥 the 512-bit minimum is giving main character energy.
5.2.3.6. Notes on Signatures 📝
okay so here's how the signing actually works, bestie. 🫶
The concatenation of the data being signed, the signature data from the version number through the hashed subpacket data (inclusive), and (for signature versions later than 3) a 6-octet trailer (see Section 5.2.4) is hashed. The resulting hash value is what is signed. The high 16 bits (first two octets) of the hash are included in the Signature packet to provide a way to reject some invalid signatures without performing a signature verification. When verifying a version 6 signature, an implementation MUST reject the signature if these octets do not match the first two octets of the computed hash. 🦋 that's a quick sanity check so you don't waste time on a bad signature, lowkey genius.
There are two fields consisting of Signature subpackets. The first field is hashed with the rest of the signature data, while the second is not hashed into the signature. The second set of subpackets (the "unhashed section") is not cryptographically protected by the signature and should include only advisory information. See Section 13.13 for more information. 🙈 the unhashed section is the group chat where anyone can type, so don't trust it blindly.
The differences between a version 4 and version 6 signature are two-fold. First, a version 6 signature increases the width of the fields that indicate the size of the hashed and unhashed subpackets, making it possible to include significantly more data in subpackets. 🌸 bigger subpacket room, literally iconic. Second, the hash is salted with random data (see Section 13.2). 🧂 salt on the hash, extra spicy. 🌶️
The algorithms for converting the hash function result to a signature are described in Section 5.2.4. 💖
5.2.3.7. Signature Subpacket Specification ✨
okay so a subpacket data set is just zero or more Signature subpackets 🎀. In Signature packets, the set gets a count in front of it: a 2-octet scalar for version 4 signatures, or a 4-octet scalar for version 6 signatures, telling you how many octets all the subpackets take up together. Add that number to a pointer and you skip right over the whole subpacket data set 🦋.
Each subpacket is a header plus a body. The header has three parts:
- The encoded subpacket length (1, 2, or 5 octets).
- The encoded Subpacket Type ID (1 octet).
- The subpacket-specific data.
The subpacket length field covers the encoded Subpacket Type ID and the subpacket-specific data, but not the length field itself 💅. It's encoded kinda like a 1-octet, 2-octet, or 5-octet OpenPGP format packet header. Here's how you decode it:
if the 1st octet < 192, then
lengthOfLength = 1
subpacketLen = 1st_octet
if the 1st octet >= 192 and < 255, then
lengthOfLength = 2
subpacketLen = ((1st_octet - 192) << 8) + (2nd_octet) + 192
if the 1st octet = 255, then
lengthOfLength = 5
subpacket length = [4-octet scalar starting at 2nd_octet]Now the critical bit, which is giving main character energy 👑. Bit 7 of the encoded Subpacket Type ID is the "critical" bit. If it's set, it means the subpacket is one the evaluator of the signature has to recognize. If a subpacket shows up with the critical bit set but the evaluating implementation doesn't know what it is, the evaluator SHOULD consider the signature to be in error. Yikes 💀.
An implementation SHOULD ignore any non-critical subpacket of a type it doesn't recognize. Easy, no drama 🌸.
An evaluator may "recognize" a subpacket but not implement it. The whole point of the critical bit is to let the signer say, "hey, I'd rather you throw an error than silently ignore this new feature I'm using" 🙈.
The other bits of the encoded Subpacket Type ID (that's bits 6-0) hold the Subpacket Type ID itself 🧚.
The following signature subpackets are defined:
+=========+===========================+==================+
| ID | Description | Reference |
+=========+===========================+==================+
| 0 | Reserved | |
+---------+---------------------------+------------------+
| 1 | Reserved | |
+---------+---------------------------+------------------+
| 2 | Signature Creation Time | Section 5.2.3.11 |
+---------+---------------------------+------------------+
| 3 | Signature Expiration Time | Section 5.2.3.18 |
+---------+---------------------------+------------------+
| 4 | Exportable Certification | Section 5.2.3.19 |
+---------+---------------------------+------------------+
| 5 | Trust Signature | Section 5.2.3.21 |
+---------+---------------------------+------------------+
| 6 | Regular Expression | Section 5.2.3.22 |
+---------+---------------------------+------------------+
| 7 | Revocable | Section 5.2.3.20 |
+---------+---------------------------+------------------+
| 8 | Reserved | |
+---------+---------------------------+------------------+
| 9 | Key Expiration Time | Section 5.2.3.13 |
+---------+---------------------------+------------------+
| 10 | Placeholder for backward | |
| | compatibility | |
+---------+---------------------------+------------------+
| 11 | Preferred Symmetric | Section 5.2.3.14 |
| | Ciphers for v1 SEIPD | |
+---------+---------------------------+------------------+
| 12 | Revocation Key | Section 5.2.3.23 |
| | (deprecated) | |
+---------+---------------------------+------------------+
| 13-15 | Reserved | |
+---------+---------------------------+------------------+
| 16 | Issuer Key ID | Section 5.2.3.12 |
+---------+---------------------------+------------------+
| 17-19 | Reserved | |
+---------+---------------------------+------------------+
| 20 | Notation Data | Section 5.2.3.24 |
+---------+---------------------------+------------------+
| 21 | Preferred Hash Algorithms | Section 5.2.3.16 |
+---------+---------------------------+------------------+
| 22 | Preferred Compression | Section 5.2.3.17 |
| | Algorithms | |
+---------+---------------------------+------------------+
| 23 | Key Server Preferences | Section 5.2.3.25 |
+---------+---------------------------+------------------+
| 24 | Preferred Key Server | Section 5.2.3.26 |
+---------+---------------------------+------------------+
| 25 | Primary User ID | Section 5.2.3.27 |
+---------+---------------------------+------------------+
| 26 | Policy URI | Section 5.2.3.28 |
+---------+---------------------------+------------------+
| 27 | Key Flags | Section 5.2.3.29 |
+---------+---------------------------+------------------+
| 28 | Signer's User ID | Section 5.2.3.30 |
+---------+---------------------------+------------------+
| 29 | Reason for Revocation | Section 5.2.3.31 |
+---------+---------------------------+------------------+
| 30 | Features | Section 5.2.3.32 |
+---------+---------------------------+------------------+
| 31 | Signature Target | Section 5.2.3.33 |
+---------+---------------------------+------------------+
| 32 | Embedded Signature | Section 5.2.3.34 |
+---------+---------------------------+------------------+
| 33 | Issuer Fingerprint | Section 5.2.3.35 |
+---------+---------------------------+------------------+
| 34 | Reserved | |
+---------+---------------------------+------------------+
| 35 | Intended Recipient | Section 5.2.3.36 |
| | Fingerprint | |
+---------+---------------------------+------------------+
| 37 | Reserved (Attested | |
| | Certifications) | |
+---------+---------------------------+------------------+
| 38 | Reserved (Key Block) | |
+---------+---------------------------+------------------+
| 39 | Preferred AEAD | Section 5.2.3.15 |
| | Ciphersuites | |
+---------+---------------------------+------------------+
| 100-110 | Private or Experimental | |
| | Use | |
+---------+---------------------------+------------------+
Table 5: OpenPGP Signature Subpacket Types RegistryNow the recommended stuff 🫶. Implementations SHOULD implement the four preferred algorithm subpackets (11, 21, 22, and 39), plus the "Features" (30) and "Reason for Revocation" (29) subpackets. To avoid surreptitious forwarding (see Section 13.12), implementations SHOULD also implement the "Intended Recipients Fingerprint" (35) subpacket. Heads up: if an implementation decides not to implement some of the preferences subpackets, it MUST default to the mandatory-to-implement algorithms so everyone can still talk to each other 🤝. And an encrypting implementation that doesn't implement the "Features" (30) subpacket SHOULD pick the encrypted data format based on the versions of the recipient keys or on external inference (see Section 13.7 for more details).
5.2.3.8. Signature Subpacket Types 🍓
A bunch of subpackets are currently defined for OpenPGP signatures 💜. Some apply to the signature itself, and some are attributes of the key. Subpackets found on a self-signature get placed on a certification the key makes about itself. Also, a key can have more than one User ID, so it can have more than one self-signature with different subpackets. Plot twist, but make sense 🌈.
A subpacket can show up in either the hashed or the unhashed subpacket sections of a signature. If a subpacket isn't hashed, the info in it can't be treated as definitive, because the cryptographic signature doesn't cover it 🙅♀️. Check out Section 13.13 for more on hashed and unhashed subpackets.
5.2.3.9. Notes on Subpackets 📝
It is totally possible for a signature to contain conflicting info in its subpackets 😭. For example, a signature might have multiple copies of a preference, or multiple expiration times. In most cases, an implementation SHOULD use the last subpacket in the hashed section of the signature, but it MAY use any conflict resolution scheme that makes more sense. Conflict resolution is intentionally left up to the implementer, bestie. Most conflicts are just syntax errors, and this loose wording lets a receiver be generous in what they accept while putting pressure on a creator to be stingy in what they generate 💅.
Some apparent conflicts actually make sense 🦋. For example, imagine a keyholder has a version 3 key and a version 4 key that share the same RSA key material. Either of those keys can verify a signature made by the other, so it may be reasonable for a signature to include an Issuer Key ID subpacket (Section 5.2.3.12) for each key, as a way of explicitly tying those keys to the signature 🔑✨.
5.2.3.10. Notes on Self-Signatures 🦋
okay so self-signatures are the ones a key makes about itself 💅 like, "hey it's me, this name belongs to my key" 🫶 there are three types: the certification ones (Type IDs 0x10-0x13), the Direct Key one (Type ID 0x1F), and the Subkey Binding one (Type ID 0x18). and the lowkey iconic part is that a cryptographically valid self-signature should be accepted from any primary key, no matter which Key Flags (Section 5.2.3.29) it has 🔥 so a primary key doesn't need 0x01 set in the first octet of its Key Flags to make a valid self-signature. no gatekeeping!! 🌈
for certification self-signatures, each User ID MAY have its own self-signature, so different subpackets can live in each one 🎀 and for Subkey Binding signatures, each subkey MUST have a self-signature 💖 subpackets in a certification self-signature apply to the User ID, subpackets in the subkey self-signature apply to the subkey, and subpackets on the Direct Key signature apply to the entire key ✨
now the preferences part is giving main character energy 👑 an implementation should read a self-signature's preference subpackets as narrowly as possible! like imagine a key with two names, Alice and Bob 🍓 Alice prefers AES-256 with OCB, and Bob prefers Camellia-256 with GCM. if the implementation finds the key through Alice's name, the preferred AEAD ciphersuite is AES-256 with OCB 🦋 but if it finds the key through Bob's name, the preferred algorithm is Camellia-256 with GCM 💜 and if the key is located by Key ID, the algorithm of the Primary User ID of the key provides the preferred AEAD ciphersuite 🧚
revoking a self-signature or letting it expire has a meaning that changes with the signature type 🙈 revoking the self-signature on a User ID basically retires that user name. the self-signature says "my name X is tied to my signing key K," and other users' certifications back it up 🫶 if another user revokes their certification, they're saying they no longer believe that name and key go together 💀 and if the users revoke their own self-signature, they no longer go by that name or use that email address, etc. 😭 revoking a binding signature retires that subkey, and revoking a Direct Key signature cancels that signature 🌸 please see Section 5.2.3.31 for more relevant details, bestie 💅
since a self-signature holds important info about how the key gets used, an implementation SHOULD allow the user to rewrite the self-signature and that important info, like preferences and key expiration ✨
when an implementation imports a secret key, it SHOULD verify that the key's internal self-signatures don't advertise features or algorithms the implementation doesn't support 🔍 if it spots a mismatch, it SHOULD warn the user and offer to create new self-signatures that advertise the actual set of features and algorithms the implementation supports 🎀
an implementation that runs into multiple self-signatures on the same object MUST select the most recent valid self-signature and ignore all the others 👑 no cap, the newest valid one wins 🔥
by convention, a version 4 key stores info about the primary Public Key (key flags, key expiration, etc.) and the Transferable Public Key as a whole (features, algorithm preferences, etc.) in a User ID self-signature of type 0x10 or 0x13 🌸 some implementations need at least one User ID with a valid self-signature present before they'll use a version 4 key 🙈 so for this reason, it is RECOMMENDED to include at least one User ID with a self-signature in version 4 keys 💖
for version 6 keys, it is RECOMMENDED to store info about the primary Public Key as well as the Transferable Public Key as a whole (key flags, key expiration, features, algorithm preferences, etc.) in a Direct Key signature (Type ID 0x1F) over the Public Key, instead of putting it in a User ID self-signature 🧚 and an implementation MUST ensure a valid Direct Key signature is present before using a version 6 key 🔒 this prevents certain attacks where an adversary strips a self-signature that specifies a Key Expiration Time or certain preferences 😭 yikes, not on my watch 💀
an implementation SHOULD NOT require a User ID self-signature to be present in order to consume or use a key, unless the particular use depends on the keyholder identifying themselves with the textual label in the User ID ✨ like when you refresh a key to learn about changes in expiration, advertised features, algorithm preferences, revocation, subkey rotation, and so forth, there's no need to require a User ID self-signature 🦋 on the other hand, when verifying a signature over an email message, an implementation MAY choose to only accept a signature from a key that has a valid self-signature over a User ID matching the message's From: header 🫶 that's a way to dodge a signature transplant attack 🛡️ periodt 💅
5.2.3.11. Signature Creation Time 🕰️
(4-octet time field)
the time the signature was made ⏰
this subpacket MUST be present in the hashed area 💖
when generating this subpacket, it SHOULD be marked as critical 🔥
5.2.3.12. Issuer Key ID 🪪
(8-octet Key ID)
the OpenPGP Key ID of the key issuing the signature 🎀 if the version of that key is greater than 4, this subpacket MUST NOT be included in the signature 🙅♀️ for these keys, consider the Issuer Fingerprint subpacket (Section 5.2.3.35) instead 🌸
Note: in previous versions of this specification, this subpacket was simply known as the "Issuer" subpacket. 💜
5.2.3.13. Key Expiration Time ⏳
(4-octet time field)
the validity period of the key 🌈 it's the number of seconds after the key creation time that the key expires 💅 for a direct or certification self-signature, the key creation time is that of the primary key 👑 for a Subkey Binding signature, the key creation time is that of the subkey 🧚 if this is not present or has a value of zero, the key never expires ✨ this is only found on a self-signature 🫶
when an implementation generates this subpacket, it SHOULD be marked as critical 🔥
5.2.3.14. Preferred Symmetric Ciphers for v1 SEIPD 🔐
(array of 1-octet values)
a series of Symmetric Cipher Algorithm IDs showing how the keyholder prefers to receive the version 1 Symmetrically Encrypted and Integrity Protected Data packet (Section 5.13.1) 🎀 the subpacket body is an ordered list of octets, most preferred first 💖 it's assumed that only the algorithms listed are supported by the recipient's implementation 🦋 algorithm IDs are defined in Section 9.3 🌸 this is only found on a self-signature 💅
when generating a v2 SEIPD packet, this preference list isn't relevant, so check out Section 5.2.3.15 instead 🧚
5.2.3.15. Preferred AEAD Ciphersuites 🌈
(array of pairs of octets indicating Symmetric Cipher and AEAD algorithms)
a series of paired algorithm IDs showing how the keyholder prefers to receive the version 2 Symmetrically Encrypted and Integrity Protected Data packet (Section 5.13.2) 💜 each pair of octets is a combo of a symmetric cipher and an AEAD mode the keyholder prefers to use 🫶 the Symmetric Cipher Algorithm ID comes before the AEAD algorithm ID in each pair 🍓 the subpacket body is an ordered list of pairs of octets, most preferred combo first 👑
it's assumed that only the combos listed are supported by the recipient's implementation, with one exception: the mandatory-to-implement combo of AES-128 and OCB 🔥 if AES-128 and OCB aren't in the subpacket, it's implicitly listed at the end 💖
AEAD algorithm IDs are listed in Section 9.6 🦋 Symmetric Cipher Algorithm IDs are listed in Section 9.3 🌸
for example, a subpacket containing the six octets
09 02 09 03 13 02means the keyholder prefers to receive v2 SEIPD using AES-256 with OCB, then AES-256 with GCM, then Camellia-256 with OCB, and finally the implicit AES-128 with OCB 🧚 iconic ✨
note that support for the version 2 Symmetrically Encrypted and Integrity Protected Data packet (Section 5.13.2) in general is indicated by a Features Flag (Section 5.2.3.32) 🎀
this subpacket is only found on a self-signature 💅
when generating a v1 SEIPD packet, this preference list isn't relevant, so see Section 5.2.3.14 instead 🙈
5.2.3.16. Preferred Hash Algorithms 🔐
(array of 1-octet values)
message digest algorithm IDs that show which algorithms the keyholder prefers to receive 💜 like the Preferred AEAD Ciphersuites, the list is ordered 🫶 algorithm IDs are defined in Section 9.5 🌈 this is only found on a self-signature ✨
5.2.3.17. Preferred Compression Algorithms 🗜️
(array of 1-octet values)
compression algorithm IDs that show which algorithms the keyholder prefers to use 🦋 like the Preferred AEAD Ciphersuites, the list is ordered 💖 algorithm IDs are defined in Section 9.4 🍓 a zero, or the absence of this subpacket, means uncompressed data is preferred; the keyholder's implementation might have no compression support available 🌸 this is only found on a self-signature 👑
5.2.3.18. Signature Expiration Time ⌛
(4-octet time field)
the validity period of the signature 🕰️ it's the number of seconds after the Signature Creation Time that the signature expires 💅 if this is not present or has a value of zero, it never expires 🧚
when an implementation generates this subpacket, it SHOULD be marked as critical 🔥
5.2.3.19. Exportable Certification 📤
(1 octet of exportability, 0 for not, 1 for exportable)
this subpacket shows whether a certification signature is "exportable," meaning it's meant for users other than the signature's issuer 🎀 the packet body holds a Boolean flag saying whether the signature is exportable 💜 if this packet isn't present, the certification is exportable; that's the same as a flag containing a 1 🌈
non-exportable, or "local," certifications are signatures a user makes to mark a key as valid within that user's implementation only 🫶
so when an implementation prepares a user's copy of a key for transport to another user (that's the process of "exporting" the key), any local certification signatures get deleted from the key 💀 the receiver of a transported key "imports" it and also trims any local certifications 🦋 in normal operation there won't be any local certifications, assuming the import is done on an exported key 🌸 but there are instances where this can reasonably happen 🙈 for example, if an implementation lets you import keys from a key database as well as an exported key, then this situation can come up 🍓
some implementations don't represent the interest of a single user (a key server, for example) 🖥️ those always trim local certifications from any key they handle 💅
when an implementation generates this subpacket and marks the signature as non-exportable, the subpacket MUST be marked as critical 🔥 no exceptions bestie 🔒
5.2.3.20. Revocable 🚫
(1 octet of revocability, 0 for not, 1 for revocable)
a Signature's revocability status 💅 the packet body holds a Boolean flag saying whether the signature is revocable 🎀 signatures that aren't revocable ignore any later Revocation Signatures 🙈 they represent the signer's commitment that its signature can't be revoked for the life of its key 💖 if this packet isn't present, the signature is revocable 🦋
5.2.3.21. Trust Signature 🤝✨
(1 octet "level" (depth), 1 octet of trust amount)
okay so this one is all about trust levels, bestie 🫶 The signer is saying the key isn't just valid, it's also trustworthy at a specific level. Level 0 is basically the same as an ordinary validity signature. Level 1 means the signed key is a valid trusted introducer, and the 2nd octet of the body tells you how much trust it gets. Level 2 means the signed key is trusted to issue level 1 Trust Signatures, so it's a "meta introducer" 👑. In general, a level n Trust Signature says the key is trusted to issue level n-1 Trust Signatures, which is giving layers 🧁. The trust amount goes from 0-255. Values under 120 mean partial trust, and 120 or higher means complete trust 💅. Implementations SHOULD emit values of 60 for partial trust and 120 for complete trust.
5.2.3.22. Regular Expression 🔍🌸
(null-terminated UTF-8 encoded Regular Expression)
This is used alongside Trust Signature packets (of level > 0) to limit how far the trust reaches 🦋 Only signatures made by the target key on User IDs that match the Regular Expression in the body of this packet get trust extended by the Trust Signature subpacket. The Regular Expression uses the same syntax as Henry Spencer's "almost public domain" Regular Expression [REGEX] package. You can read a description of the syntax in Section 8. It matches (or does not match) a sequence of UTF-8-encoded Unicode characters from User IDs, and the expression itself is also written with UTF-8 characters.
Okay, the historical drama 🙈 This subpacket includes a null character (an octet with value zero) after the Regular Expression, for historical reasons. When an implementation parses a Regular Expression subpacket, it MUST remove this octet. If the octet isn't there, it MUST reject the subpacket. That means ignoring it if it's non-critical and rejecting the signature if it's critical. When an implementation generates a Regular Expression subpacket, it MUST include the null terminator.
When generating this subpacket, it SHOULD be marked as critical.
5.2.3.23. Revocation Key (Deprecated) 🚫💀
(1 octet of class, 1 octet of public key algorithm ID, 20 octets of version 4 fingerprint)
Bestie, this one is deprecated, so it's a no-go ❌ Applications MUST NOT generate such a subpacket.
If you want the delegated revocation vibe, use an escrowed Revocation Signature instead. See Section 13.9 for more details ✨
The rest of this section describes how some implementations try to interpret this deprecated subpacket, so here we go 🧐
It was meant to authorize the specified key to issue Revocation Signatures for this key. The class octet must have bit 0x80 set. If bit 0x40 is set, the revocation information is sensitive. The other bits are saved for future expansion to other kinds of authorizations. This is only found on a Direct Key self-signature (Type ID 0x1F). Its use on other types of self-signatures is unspecified.
If the "sensitive" flag is set, the keyholder is saying this subpacket holds private trust information about a real-world sensitive relationship 🤫 In that case, implementations SHOULD NOT export this signature to other users, except when the data needs to be available. That means when the signature is being sent to the designated revoker, or when it comes with a Revocation Signature from that revoker. It may be appropriate to isolate this subpacket within a separate signature so that it isn't combined with other subpackets that need to be exported 🧚
5.2.3.24. Notation Data 📝💜
(4 octets of flags, 2 octets of name length (M), 2 octets of value length (N), M octets of name data, N octets of value data)
This subpacket describes a "notation" that the issuer of the signature wants to attach 🎀 A notation has a name and a value, and each is a string of octets. A signature can have more than one notation, and notations can be used for any extension the issuer cares about. The "flags" field holds 4 octets of flags.
All undefined flags MUST be zero. The defined flags look like this:
+=======================+================+================+
| Flag Position | Shorthand | Description |
+=======================+================+================+
| 0x80000000 (first bit | human-readable | Notation value |
| of the first octet) | | is UTF-8 text |
+-----------------------+----------------+----------------+
Table 6: OpenPGP Signature Notation Data Subpacket
Notation Flags RegistryNotation names are arbitrary strings encoded in UTF-8, and they live in two namespaces: the IETF namespace and the user namespace 🌈
The IETF namespace is registered with IANA. These names MUST NOT contain the "@" character (0x40). That's the tag for the user namespace, so it's a big no 🙅♀️
+===============+===========+================+
| Notation Name | Data Type | Allowed Values |
+===============+===========+================+
| No registrations at this time. |
+============================================+
Table 7: OpenPGP Signature Notation Data
Subpacket Types RegistryThis registry is initially empty 🫧
Names in the user namespace are a UTF-8 string tag, then "@", then a DNS domain name. The tag MUST NOT contain an "@" character. For example, the "sample" tag used by Example Corporation could be sample@example.com 💖
Names in a user space are owned and controlled by the owners of that domain. Obviously, it's bad form to make a new name in a DNS space you don't own 😭
Since the user namespace is in the form of an email address, implementers MAY wish to arrange for that address to reach a person who can be consulted about the use of the named tag. Note that because of UTF-8 encoding, not every valid user space name tag is a valid email address 📧
If there is a critical notation, the criticality applies to that specific notation and not to notations in general ✨
5.2.3.25. Key Server Preferences 🗄️💅
(N octets of flags)
This is a list of 1-bit flags showing the keyholder's preferences for how the key is handled on a key server. All undefined flags MUST be zero.
+=========+===========+===========================================+
| Flag | Shorthand | Definition |
+=========+===========+===========================================+
| 0x80... | No-modify | The keyholder requests that this key only |
| | | be modified or updated by the keyholder |
| | | or an administrator of the key server. |
+---------+-----------+-------------------------------------------+
Table 8: OpenPGP Key Server Preference Flags RegistryThis only shows up on a self-signature 🔐
5.2.3.26. Preferred Key Server 🌐🦋
(String)
This is a URI for a key server that the keyholder would like used for updates 💜 Keys with multiple User IDs can have a Preferred Key Server for each User ID, so you can get really specific. Also, since this is a URI, the key server can actually be a copy of the key retrieved over https, ftp, http, and so on 🍓
5.2.3.27. Primary User ID 👑🎀
(1 octet, Boolean)
This is a flag in a User ID's self-signature that says whether this User ID is the main one for this key 💖 It's reasonable for an implementation to settle preference ambiguities by looking at the Primary User ID, for example. If this flag is absent, its value is zero. If more than one User ID in a key is marked as primary, the implementation may resolve the ambiguity in any way it sees fit. It is RECOMMENDED that priority go to the User ID with the most recent self-signature ⭐
When this shows up on a self-signature on a User ID packet, it only applies to User ID packets. When it shows up on a self-signature on a User Attribute packet, it only applies to User Attribute packets. So there are two different, independent "primaries": one for User IDs and one for User Attributes. Two primaries, zero drama 🫶
5.2.3.28. Policy URI 📜✨
(String)
This subpacket holds a URI for a document that describes the policy under which the signature was issued 🌸
5.2.3.29. Key Flags ✨
(N octets of flags)
okay so this subpacket is basically a little list of binary flags that tell you what a key is for 💅 It's a string of octets, and an implementation MUST NOT assume a fixed size, so it can grow over time 🌱 If a list is shorter than an implementation expects, the flags that aren't mentioned are considered to be zero 🫶 The defined flags look like this:
+===========+======================================================+
| Flag | Definition |
+===========+======================================================+
| 0x01... | This key may be used to make User ID certifications |
| | (Signature Type IDs 0x10-0x13) or Direct Key |
| | signatures (Signature Type ID 0x1F) over other keys. |
+-----------+------------------------------------------------------+
| 0x02... | This key may be used to sign data. |
+-----------+------------------------------------------------------+
| 0x04... | This key may be used to encrypt communications. |
+-----------+------------------------------------------------------+
| 0x08... | This key may be used to encrypt storage. |
+-----------+------------------------------------------------------+
| 0x10... | The private component of this key may have been |
| | split by a secret-sharing mechanism. |
+-----------+------------------------------------------------------+
| 0x20... | This key may be used for authentication. |
+-----------+------------------------------------------------------+
| 0x80... | The private component of this key may be in the |
| | possession of more than one person. |
+-----------+------------------------------------------------------+
| 0x0004... | Reserved (ADSK) |
+-----------+------------------------------------------------------+
| 0x0008... | Reserved (timestamping) |
+-----------+------------------------------------------------------+
Table 9: OpenPGP Key Flags RegistryUsage notes:
The flags in this packet may show up in self-signatures or in certification signatures 🎀 They mean different things depending on who's making the statement, which is so relevant 🙈 For example, a certification signature that has the "sign data" flag is saying the certification is for that use. But the "communications encryption" flag in a self-signature is stating a preference that a given key be used for communications. Also, figuring out what counts as "communications" versus "storage" is a seriously thorny issue 😭 That decision is left totally up to the implementation. The authors of this document don't claim any special wisdom on it, and they realize accepted opinion may change 🌈
The "split key" (0x10) and "group key" (0x80) flags go on a self-signature only. They're meaningless on a certification signature 🚫 They SHOULD be placed only on a Direct Key signature (Type ID 0x1F) or a Subkey Binding signature (Type ID 0x18), one that refers to the key the flag applies to 🧚
When an implementation generates this subpacket, it SHOULD be marked as critical 👑
5.2.3.30. Signer's User ID ✨
(String)
This subpacket lets a keyholder say which User ID is responsible for the signing 💜 Lots of keyholders use one key for different vibes, like business communications and personal ones. This subpacket is how they make it clear which of their roles is making a signature 💼
This subpacket is not appropriate to use to refer to a User Attribute packet 🙅♀️
5.2.3.31. Reason for Revocation ✨
(1 octet of revocation code, N octets of reason string)
This subpacket is only used in Key Revocation and Certification Revocation signatures 🦋 It describes why the key or certification got revoked 💀
The first octet holds a machine-readable code for the reason:
+=========+========================================+
| Code | Reason |
+=========+========================================+
| 0 | No reason specified (Key Revocation or |
| | Certification Revocation signatures) |
+---------+----------------------------------------+
| 1 | Key is superseded (Key Revocation |
| | signatures) |
+---------+----------------------------------------+
| 2 | Key material has been compromised (Key |
| | Revocation signatures) |
+---------+----------------------------------------+
| 3 | Key is retired and no longer used (Key |
| | Revocation signatures) |
+---------+----------------------------------------+
| 32 | User ID information is no longer valid |
| | (Certification Revocation signatures) |
+---------+----------------------------------------+
| 100-110 | Private Use |
+---------+----------------------------------------+
Table 10: OpenPGP Reason for Revocation
(Revocation Octet) RegistryAfter the code comes a string of octets that explains the Reason for Revocation in human-readable form (UTF-8) 📝 The string may be null (zero length) 🫧 The length of the subpacket is the length of the reason string plus one. An implementation SHOULD implement this subpacket, include it in all Revocation Signatures, and interpret revocations appropriately 🔥 The reasons have important semantic differences, which is exactly why there are important reasons to revoke signatures at all 😤
If a key got revoked because of a compromise, then all signatures created by that key are suspect 🚨 But if it was merely superseded or retired, old signatures are still valid 💅 If the revoked signature is the self-signature certifying a User ID, the revocation means that user name is no longer in use 🪦 Such a signature revocation SHOULD include a Reason for Revocation subpacket containing code 32 ✨
Also, any certification may be revoked, including one on somebody else's key 👀 There are lots of good reasons to revoke a certification signature, like when a keyholder leaves the employ of a business that has an email address 💌 A revoked certification is no longer part of validity calculations, so it's gone from the math 🧮
5.2.3.32. Features ✨
(N octets of flags)
The Features subpacket shows which advanced OpenPGP features a user's implementation supports 🌸 Here's the thing: when features get added to OpenPGP that can't be backward compatible, this lets a user say "hey, I can use that one" 🙌 The flags are single bits that indicate a given feature is supported.
This subpacket is kind of like a preferences subpacket, and it only shows up in a self-signature 🪞
An implementation SHOULD NOT use a feature listed when sending to a user who doesn't say they can use it, unless the implementation can infer support for the feature from another implementation-dependent mechanism 🙈 Translation: don't just assume, bestie 🚫
Defined features are as follows:
First octet:
+=========+=====================================+===========+
| Feature | Definition | Reference |
+=========+=====================================+===========+
| 0x01... | Version 1 Symmetrically Encrypted | Section |
| | and Integrity Protected Data packet | 5.13.1 |
+---------+-------------------------------------+-----------+
| 0x02... | Reserved | |
+---------+-------------------------------------+-----------+
| 0x04... | Reserved | |
+---------+-------------------------------------+-----------+
| 0x08... | Version 2 Symmetrically Encrypted | Section |
| | and Integrity Protected Data packet | 5.13.2 |
+---------+-------------------------------------+-----------+
Table 11: OpenPGP Features Flags RegistryIf an implementation implements any of the defined features, it SHOULD implement the Features subpacket too 🫶
See Section 13.7 for details on how to use the Features subpacket when generating encryption data 🔐
5.2.3.33. Signature Target ✨
(1 octet public key algorithm, 1 octet hash algorithm, N octets hash)
This subpacket identifies a specific target signature that a signature refers to 🎯 For Revocation Signatures, it gives explicit designation of which signature is being revoked. For a Third-Party Confirmation or Timestamp signature, it designates which signature gets signed. All the arguments are an identifier of that target signature 🪪
The N octets of hash data MUST be the size of the signature's hash 📏 For example, a target signature with a SHA-1 hash MUST have 20 octets of hash data. No wiggle room there, babe 💖
5.2.3.34. Embedded Signature ✨
(1 Signature packet body)
This subpacket contains a complete Signature packet body as specified in Section 5.2 📦 It's useful when one signature needs to refer to another signature, or be incorporated into it 🪆
5.2.3.35. Issuer Fingerprint ✨
(1 octet key version number, N octets of fingerprint)
This is the OpenPGP Key fingerprint of the key that issued the signature 🪪 This subpacket SHOULD be included in all signatures. If the version of the issuing key is 4 and an Issuer Key ID subpacket (Section 5.2.3.12) is also included in the signature, the Key ID of the Issuer Key ID subpacket MUST match the low 64 bits of the fingerprint 🔗
Note that the length N of the fingerprint for a version 4 key is 20 octets. For a version 6 key, N is 32 📐 Since the version of the signature is bound to the version of the key, the version octet here MUST match the version of the signature. If the version octet does not match the signature version, the receiving implementation MUST treat it as a malformed signature (see Section 5.2.5) 🚩 Not a vibe, not valid, bye 👋
5.2.3.36. Intended Recipient Fingerprint ✨
(1 octet key version number, N octets of fingerprint)
This is the OpenPGP Key fingerprint of the intended recipient primary key 🎀 If one or more subpackets of this type are included in a signature, it SHOULD be considered valid only in an encrypted context, where the key it was encrypted to is one of the indicated primary keys or one of their subkeys 🔒 This is how you stop someone from forwarding a signature outside of its intended, encrypted context (see Section 13.12) 🛡️
Note that the length N of the fingerprint for a version 4 key is 20 octets. For a version 6 key, N is 32 📏
An implementation SHOULD generate this subpacket when creating a signed and encrypted message 💌
When generating this subpacket in a version 6 signature, it SHOULD be marked as critical 👑
5.2.4. Computing Signatures ✨
Okay bestie, every signature works the same way: you hash up all the signature data into one blob, then feed that hash into the signature algorithm 🔐 It's giving "hash first, sign second" and it's honestly iconic 💅
When you're making or verifying a version 6 signature, the salt gets fed into the hash context before any other data 🧂
For binary document signatures (Type ID 0x00), the document data is hashed directly 📄 For text document signatures (Type ID 0x01), the implementation MUST first canonicalize the document by converting line endings to <CR><LF> and encoding it in UTF-8 (see [RFC3629]) 🌸 Then the resulting UTF-8 byte stream gets hashed.
Now for signatures over keys 🔑 For a version 4 signature over a key, the hash data starts with the octet 0x99, then a 2-octet length of the key, then the body of the key packet. For a version 6 signature over a key, it starts with the salt, then the octet 0x9B, then a 4-octet length of the key, then the body of the key packet.
A Subkey Binding signature (Type ID 0x18) or Primary Key Binding signature (Type ID 0x19) then hashes the subkey using the same format as the main key, so it also starts with 0x99 or 0x9B as the first octet 🦋 A Primary Key Revocation signature (Type ID 0x20) only hashes the key being revoked 🚫 A Subkey Revocation signature (Type ID 0x28) hashes the primary key first, and then the subkey being revoked.
Certification signatures (Type IDs 0x10 through 0x13) hash the User ID bound to the key into the hash context after all of the above 🎀 A version 3 certification hashes the contents of the User ID or User Attribute packet without the packet header. A version 4 or version 6 certification hashes the constant 0xB4 for User ID certifications or the constant 0xD1 for User Attribute certifications. After that comes a 4-octet number giving the length of the User ID or User Attribute data, and then the data itself 🧚
Third-Party Confirmation signatures (Type ID 0x50) are a whole mood 🙈 They hash the salt (version 6 signatures only), then the octet 0x88, then the 4-octet length of the signature, and then the body of the Signature packet. (Note that this is a Legacy packet header for a Signature packet with the length-of-length field set to zero.) The unhashed subpacket data of the Signature packet being hashed is left out of the hash, and the unhashed subpacket data length value is set to zero.
Once the data body is hashed, a trailer gets hashed too ✨ The trailer depends on the signature version.
- A version 3 signature hashes five octets of the packet body, starting from the signature type field. That's the signature type, followed by the 4-octet Signature Creation Time.
- A version 4 or version 6 signature hashes the packet body starting from its first field, the version number, all the way through the end of the hashed subpacket data and a final extra trailer 👑 So the hashed fields are:
- an octet indicating the signature version (
0x04for version 4, and0x06for version 6), - the signature type,
- the public key algorithm,
- the hash algorithm,
- the hashed subpacket length,
- the hashed subpacket body,
- a second version octet (
0x04for version 4, and0x06for version 6), - a single octet
0xFF, and - a number representing the length (in octets) of the hashed data from the Signature packet through the hashed subpacket body. This is a 4-octet big-endian unsigned integer of the length modulo
2^32.
- an octet indicating the signature version (
After all of this has been hashed in a single hash context, the resulting hash field gets used in the signature algorithm. Its first two octets go into the Signature packet, as described in Section 5.2.3 🌈
For worked examples of the data hashed during a signature, check out Appendix A.3.1 📚
5.2.4.1. Notes about Signature Computation
Here's the tea ☕ The data OpenPGP actually hashes changes depending on the signature version. This makes sure a signature made with one version can't get repurposed as a signature with a different version over subtly different data 🕵️♀️ The hashed data streams differ based on their trailer, most critically in the fifth and sixth octets from the end of the stream. In particular:
- A version 3 signature uses the fifth octet from the end to store its Signature Type ID. This MUST NOT be Signature Type ID
0xFF. - All signature versions later than version 3 always use a literal
0xFFin the fifth octet from the end. For these later versions, the sixth octet from the end (the octet before the0xFF) stores the signature version number.
5.2.5. Malformed and Unknown Signatures
Sometimes a Signature packet (or its matching One-Pass Signature packet; see Section 5.4) is malformed or unknown 💀 Some examples of the problems it might run into are below. This list is not exhaustive, so there could be more chaos out there 🙃
- An unknown signature type
- An unknown signature version
- An unsupported signature version
- An unknown "critical" subpacket (see Section 5.2.3.7) in the hashed area
- A subpacket with a length that diverges from the expected length
- A hashed subpacket area with a length that exceeds the length of the Signature packet itself
- A hash algorithm known to be weak (e.g., MD5)
- A mismatch between the expected salt length of the hash algorithm and the actual salt length
- A mismatch between the One-Pass Signature version and the Signature version (see Section 10.3.2.2)
- A signature with a version other than 6, made by a version 6 key
When an implementation runs into one of these malformed or unknown signatures, it MUST ignore the signature for validation purposes. It MUST NOT indicate a successful signature validation for that signature 🚫 At the same time, it MUST NOT halt processing on the packet stream or reject other signatures in the same packet stream just because an unknown or invalid signature shows up 💖
This is a big deal for forward compatibility! Giving back an output that says no successful signatures were found is way better than aborting everything. Please don't be the parser that rage-quits 😭
5.3. Symmetric Key Encrypted Session Key Packet (Type ID 3)
The Symmetric Key Encrypted Session Key (SKESK) packet holds the symmetric key encryption of a session key used to encrypt a message 🔒 Zero or more Public Key Encrypted Session Key packets (Section 5.1) and/or Symmetric Key Encrypted Session Key packets come before an encryption container (that is, a Symmetrically Encrypted and Integrity Protected Data packet or -- for historic data -- a Symmetrically Encrypted Data packet) that holds an Encrypted Message. The message is encrypted with a session key, and that session key is itself encrypted and stored in the Encrypted Session Key packet(s). Bestie, it's a key inside a key inside a message 🪆
If the encryption container is preceded by one or more Symmetric Key Encrypted Session Key packets, each one specifies a passphrase that may be used to decrypt the message 🗝️ This lets you encrypt a message to a bunch of public keys, and also to one or more passphrases. Main character energy ✨
The body of this packet starts with a 1-octet number giving the version number of the packet type. The currently defined versions are 4 and 6. The rest of the packet depends on the version 🎀
The versions differ in how they encrypt the session key with the passphrase, and in what they encode. The version of the SKESK packet must align with the version of the SEIPD packet (see Section 10.3.2.1). Any new version of the SKESK packet should be registered in the registry established in Section 10.3.2.1 📋
5.3.1. Version 4 Symmetric Key Encrypted Session Key Packet Format
A v4 SKESK packet precedes a v1 SEIPD (see Section 5.13.1). In historic data, it's sometimes found preceding a deprecated SED packet (see Section 5.7). A v4 SKESK packet MUST NOT precede a v2 SEIPD packet (see Section 10.3.2.1) 🙅♀️
A version 4 Symmetric Key Encrypted Session Key packet consists of:
- A 1-octet version number with value 4.
- A 1-octet number describing the symmetric algorithm used.
- An S2K Specifier. Its length depends on its type (see Section 3.7.1).
- Optionally, the encrypted session key itself, which is decrypted with the S2K object.
If the encrypted session key is not present (you can tell by the packet length and the S2K Specifier size), then the S2K algorithm applied to the passphrase produces the session key for decrypting the message. It uses the Symmetric Cipher Algorithm ID from the Symmetric Key Encrypted Session Key packet 🧁
If the encrypted session key is present, the result of applying the S2K algorithm to the passphrase is used to decrypt just that encrypted session key field, using CFB mode with an IV of all zeros. The decryption result is a 1-octet algorithm identifier that specifies the symmetric key encryption algorithm used on the following encryption container, followed by the session key octets themselves 🌷
Note: because an all-zero IV is used for this decryption, the S2K Specifier MUST use a salt value, a Salted S2K, an Iterated and Salted S2K, or Argon2. The salt value makes sure the decryption key isn't repeated even if the passphrase gets reused 🧂
5.3.2. Version 6 Symmetric Key Encrypted Session Key Packet Format
A v6 SKESK packet precedes a v2 SEIPD packet (see Section 5.13.2). A v6 SKESK packet MUST NOT precede a v1 SEIPD packet or a deprecated Symmetrically Encrypted Data packet (see Section 10.3.2.1) 🚫
A version 6 Symmetric Key Encrypted Session Key packet consists of:
- A 1-octet version number with value 6.
- A 1-octet scalar octet count for the 5 fields following this octet.
- A 1-octet Symmetric Cipher Algorithm ID (from Table 21).
- A 1-octet AEAD algorithm identifier (from Table 25).
- A 1-octet scalar octet count of the following field.
- An S2K Specifier. Its length depends on its type (see Section 3.7.1).
- A starting IV of the size specified by the AEAD algorithm.
- The encrypted session key itself.
- An authentication tag for the AEAD mode.
A key-encryption key (KEK) is derived using HKDF [RFC5869] with SHA256 [RFC6234] as the hash algorithm 🦄 The Initial Keying Material (IKM) for HKDF is the key derived from S2K. No salt is used. The info parameter is made up of the Packet Type ID in OpenPGP format encoding (bits 7 and 6 are set, and bits 5-0 carry the Packet Type ID), the packet version, and the cipher-algo and AEAD-mode used to encrypt the key material.
Then, the session key is encrypted using the resulting key, with the AEAD algorithm specified for the version 2 Symmetrically Encrypted and Integrity Protected Data packet. Note that no chunks are used and there's only one authentication tag 🫶 The Packet Type ID encoded in OpenPGP format (bits 7 and 6 are set, and bits 5-0 carry the Packet Type ID), the packet version number, the cipher algorithm ID, and the AEAD algorithm ID are given as additional data. For example, the additional data used with AES-128 with OCB consists of the octets 0xC3, 0x06, 0x07, and 0x02 🌈
5.4. One-Pass Signature Packet (Type ID 4) 🌸
Okay bestie, the One-Pass Signature packet shows up before the signed data 🦋 It's giving "heads up, here's what's coming" energy ✨ It carries enough info that the receiver can start working out any hashes it needs to verify the signature. That's the glow-up: the actual Signature packet gets to chill at the end of the message, so the signer can compute the whole signed message in one pass 💅
The body of this packet is made of:
- A 1-octet version number. The currently defined versions are 3 and 6. Any new One-Pass Signature packet version should be registered in the registry established in Section 10.3.2.2. 💖
- A 1-octet Signature Type ID. Signature types are described in Section 5.2.1. 🎀
- A 1-octet number describing the hash algorithm used. 🍓
- A 1-octet number describing the public key algorithm used. 🌈
- Only for version 6 packets, a variable-length field containing:
- A 1-octet salt size. The value MUST match the value defined for the hash algorithm as specified in Table 23. 🫶
- The salt, which is a random value of the specified size. The value MUST match the salt field of the corresponding Signature packet. 🔥
- Only for v3 packets, an 8-octet number holding the Key ID of the signing key. 🧚
- Only for version 6 packets, 32 octets of the fingerprint of the signing key. Since a version 6 signature can only be made by a version 6 key, the length of the fingerprint is fixed. 👑
- A 1-octet number holding a flag showing whether the signature is nested. A zero value indicates that the next packet is another One-Pass Signature packet that describes another signature to be applied to the same message data. 🦋
When generating a one-pass signature, the OPS packet version MUST correspond to the version of the associated Signature packet. The historical accident is that version 4 keys use a version 3 One-Pass Signature packet, since there is no version 4 OPS 🙈 See Section 10.3.2.2 for the full correspondence of versions between Keys, Signatures, and One-Pass Signatures.
Note that if a message contains more than one one-pass signature, the Signature packets bracket the message. So the first Signature packet after the message corresponds to the last One-Pass Signature packet, and the final Signature packet corresponds to the first One-Pass Signature packet. It's like a little mirror setup 🪞✨
5.5. Key Material Packets 🔑
A key material packet holds all the info about a public or private key. There are four variants of this packet type: two major versions (versions 4 and 6) and two strongly deprecated versions (versions 2 and 3). So yes, this section is complex, and I'm not going to pretend otherwise 😭
For historical reasons, versions 1 and 5 of the key packets are unspecified. Not everything gets a seat at the table 🙈
5.5.1. Key Packet Variants 🌸
5.5.1.1. Public Key Packet (Type ID 6)
A Public Key packet starts a series of packets that forms an OpenPGP Key, which is sometimes called an OpenPGP certificate. It's the main character that kicks off the whole key bundle 👑
5.5.1.2. Public Subkey Packet (Type ID 14)
A Public Subkey packet (Type ID 14) has exactly the same format as a Public Key packet, but it denotes a subkey. One or more subkeys may be associated with a top-level key. By convention, the top-level key offers certification capability but doesn't provide encryption services, while a dedicated subkey provides encryption (see Section 10.1.5). Divide and conquer, bestie 💅
5.5.1.3. Secret Key Packet (Type ID 5)
A Secret Key packet contains all the information found in a Public Key packet, including the public key material. It also includes the secret key material after all the public key fields. Secrets, but make it structured 🤫🔐
5.5.1.4. Secret Subkey Packet (Type ID 7)
A Secret Subkey packet (Type ID 7) is the subkey analog of the Secret Key packet, and it has exactly the same format. Same vibe, just one level down 🦋
5.5.2. Public Key Packet Formats 🎀
There are four versions of key material packets. Versions 2 and 3 have been deprecated since 1998, and version 4 has been deprecated by this document. Ngl, it's a lot of retirement parties 💀
OpenPGP implementations SHOULD create keys with version 6 format. Version 4 keys are deprecated, so an implementation SHOULD NOT generate a version 4 key, but SHOULD accept it. Version 3 keys are deprecated, so an implementation MUST NOT generate a version 3 key, but MAY accept it. Version 2 keys are deprecated too, so an implementation MUST NOT generate a version 2 key, but MAY accept it. Basically: make new ones in v6, be nice to the old ones 🫶
Any new Key Version must be registered in the registry established in Section 10.3.2.2. Rules are rules, babe 📋
5.5.2.1. Version 3 Public Keys
Version 2 keys are identical to version 3 keys except for the version number. A version 3 Public Key or Public Subkey packet contains:
- A 1-octet version number (3).
- A 4-octet number denoting the time that the key was created.
- A 2-octet number denoting the time in days that the key is valid. If this number is zero, then it does not expire.
- A 1-octet number denoting the public key algorithm of the key.
- A series of multiprecision integers comprising the key material:
- MPI of RSA public modulus
n. - MPI of RSA public encryption exponent
e.
- MPI of RSA public modulus
Version 3 keys are deprecated, and they come with three weaknesses 🙃 First, it's relatively easy to construct a version 3 key that has the same Key ID as any other key, because the Key ID is simply the low 64 bits of the public modulus. Second, the fingerprint of a version 3 key hashes the key material but not its length, so there's an increased opportunity for fingerprint collisions. Third, there are weaknesses in the MD5 hash algorithm that cause developers to prefer other algorithms. See Section 5.5.4 for a fuller discussion of Key IDs and fingerprints. Not the vibe we want, periodt 💀
5.5.2.2. Version 4 Public Keys
The version 4 format is similar to the version 3 format, except for the absence of a validity period. That's been moved to the Signature packet, so it's giving "organized" 📅 In addition, fingerprints of version 4 keys are calculated differently from version 3 keys, as described in Section 5.5.4.
A version 4 packet contains:
- A 1-octet version number (4).
- A 4-octet number denoting the time that the key was created.
- A 1-octet number denoting the public key algorithm of the key.
- A series of values comprising the key material. This is algorithm specific and described in Section 5.5.5. 🔑
5.5.2.3. Version 6 Public Keys
The version 6 format is similar to the version 4 format, except for the addition of a count for the key material. This count helps parsing Secret Key packets (which are an extension of the Public Key packet format) in the case of an unknown algorithm. That's the lowkey genius part 🧠✨ In addition, fingerprints of version 6 keys are calculated differently from version 4 keys, as described in Section 5.5.4.
A version 6 packet contains:
- A 1-octet version number (6).
- A 4-octet number denoting the time that the key was created.
- A 1-octet number denoting the public key algorithm of the key.
- A 4-octet scalar octet count for the public key material specified in the next field.
- A series of values comprising the public key material. This is algorithm specific and described in Section 5.5.5. 👑🌈
5.5.3. Secret Key Packet Formats ✨
okay bestie, the Secret Key and Secret Subkey packets are basically the public key packets plus a secret stash 🔐. They have all the same data as a Public Key or Public Subkey packet, and then they tack on algorithm-specific secret key data, usually encrypted 💅
The packet contains:
- The fields of a Public Key or Public Subkey packet, as described above.
- One octet (the "S2K usage octet") that says whether and how the secret key material is protected by a passphrase 🎀. Zero means the secret key data is not encrypted. 253 (AEAD), 254 (CFB), or 255 (MalleableCFB) means an S2K Specifier and other parameters come next. Any other value is a symmetric key encryption algorithm identifier. A version 6 packet MUST NOT use the value 255 (MalleableCFB) 🙅♀️.
- Only for a version 6 packet where the secret key material is encrypted (that is, where the previous octet is not zero), a 1-octet scalar octet count of the cumulative length of all the following conditionally included S2K parameter fields 🧮.
- Conditionally included S2K parameter fields:
- If the S2K usage octet was 253, 254, or 255, a 1-octet symmetric key encryption algorithm.
- If the S2K usage octet was 253 (AEAD), a 1-octet AEAD algorithm.
- Only for a version 6 packet, and if the S2K usage octet was 253 or 254, a 1-octet count of the size of the one field following this octet.
- If the S2K usage octet was 253, 254, or 255, an S2K Specifier. The length of the S2K Specifier depends on its type (see Section 3.7.1).
- If the S2K usage octet was 253 (AEAD), an IV of a size specified by the AEAD algorithm (see Section 5.13.2), which is used as the nonce for the AEAD algorithm.
- If the S2K usage octet was 254, 255, or a cipher algorithm ID (that is, the secret data uses some form of CFB encryption), an IV of the same length as the cipher's block size.
- Plain or encrypted multiprecision integers holding the secret key data 🔑. This is algorithm specific and described in Section 5.5.5. If the S2K usage octet is 253 (AEAD), then an AEAD authentication tag sits at the end of that data. If the S2K usage octet is 254 (CFB), a 20-octet SHA-1 hash of the plaintext of the algorithm-specific portion is appended to the plaintext and encrypted with it. If the S2K usage octet is 255 (MalleableCFB) or another non-zero value (that is, a symmetric key encryption algorithm identifier), a 2-octet checksum of the plaintext of the algorithm-specific portion (sum of all octets, mod 65536) is appended to the plaintext and encrypted with it. (This is deprecated and SHOULD NOT be used; see below. 🥲)
- Only for a version 3 or 4 packet where the S2K usage octet is zero, a 2-octet checksum of the algorithm-specific portion (sum of all octets, mod 65536).
The details about storing algorithm-specific secrets above are summarized in Table 2 📋.
Note that the version 6 packet format adds two count values to help parsing packets with unknown S2K or public key algorithms 🧠.
Secret MPI values can be encrypted using a passphrase 🤫. If an S2K Specifier is given, it describes the algorithm for converting the passphrase to a key. Otherwise, a simple MD5 hash of the passphrase is used. An implementation producing a passphrase-protected Secret Key packet MUST use an S2K Specifier. The simple hash is only for read-only backward compatibility, though implementations MAY continue to use existing private keys in the old format 🕰️. The cipher for encrypting the MPIs is specified in the Secret Key packet.
Encryption and decryption of the secret data use the key created from the passphrase and the IV from the packet. If the S2K usage octet is not 253, CFB mode is used. Version 3 keys (which are only RSA) use a different mode than other key formats. With version 3 keys, the MPI bit count prefix (that is, the first two octets) is not encrypted. Only the MPI non-prefix data is encrypted. On top of that, the CFB state is resynchronized at the beginning of each new MPI value so that the CFB block boundary lines up with the start of the MPI data 🧩.
With version 4 and version 6 keys, there's a simpler method. All secret MPI values are encrypted, including the MPI bit count prefix ✨.
If the S2K usage octet is 253, the KEK is derived using HKDF [RFC5869] to provide key separation. SHA256 [RFC6234] is the hash algorithm for HKDF. The IKM for HKDF is the key derived from S2K. No salt is used. The info parameter is made of the Packet Type ID encoded in OpenPGP format (bits 7 and 6 are set, and bits 5-0 carry the Packet Type ID), the packet version, and the cipher-algo and AEAD-mode used to encrypt the key material 🦋.
Then the encrypted MPI values get encrypted as one combined plaintext using one of the AEAD algorithms specified for the version 2 Symmetrically Encrypted and Integrity Protected Data packet. Note that no chunks are used and there is only one authentication tag 🔒. As additional data, the Packet Type ID in OpenPGP format encoding (bits 7 and 6 are set, and bits 5-0 carry the Packet Type ID), followed by the Public Key packet fields, starting with the packet version number, are passed to the AEAD algorithm. For example, the additional data used with a Secret Key packet of version 4 consists of the octets 0xC5, 0x04, followed by four octets of creation time, one octet denoting the public key algorithm, and the algorithm-specific public key parameters. For a Secret Subkey packet, the first octet would be 0xC7. For a version 6 key packet, the second octet would be 0x06, and the 4-octet octet count of the public key material would be included as well (see Section 5.5.2) 🧁.
The 2-octet checksum that follows the algorithm-specific portion is the algebraic sum, mod 65536, of the plaintext of all the algorithm-specific octets (including the MPI prefix and data). With version 3 keys, the checksum is stored in the clear. With version 4 keys, the checksum is encrypted like the algorithm-specific data. This value is used to check that the passphrase was correct. However, this checksum is deprecated, and an implementation SHOULD NOT use it 🙅♀️. Instead, an implementation should use the SHA-1 hash denoted with a usage octet of 254. The reason is that there are some attacks that involve modifying the secret key undetected 😱. If the S2K usage octet is 253, no checksum or SHA-1 hash is used, but the authentication tag of the AEAD algorithm follows instead.
When decrypting the secret key material using any of these schemes (that is, where the usage octet is non-zero), the resulting cleartext octet stream must be well formed. Be careful here, bestie: an implementation MUST NOT interpret octets beyond the unwrapped cleartext octet stream as part of any of the unwrapped MPI objects. An implementation MUST also reject any secret key material whose cleartext length does not align with the lengths of the unwrapped MPI objects as unusable 🚫.
5.5.4. Key IDs and Fingerprints 🪪
Every OpenPGP Key has a fingerprint and a Key ID. The way these values get computed depends on the key version. The fingerprint length changes with the key version, but the Key ID (which is only used in v3 PKESK packets; see Section 5.1.1) is always 64 bits. The following registry represents the subsections below 👀:
+=======+===================+===============+=============+=========+
|Key | Fingerprint | Fingerprint | Key ID |Reference|
|Version| | Length | | |
| | | (Bits) | | |
+=======+===================+===============+=============+=========+
|3 | MD5(MPIs without | 128 | low 64 bits |Section |
| | length octets) | | of RSA |5.5.4.1 |
| | | | modulus | |
+-------+-------------------+---------------+-------------+---------+
|4 | SHA1(normalized | 160 | last 64 |Section |
| | pubkey packet) | | bits of |5.5.4.2 |
| | | | fingerprint | |
+-------+-------------------+---------------+-------------+---------+
|6 | SHA256(normalized | 256 | first 64 |Section |
| | pubkey packet) | | bits of |5.5.4.3 |
| | | | fingerprint | |
+-------+-------------------+---------------+-------------+---------+
Table 12: OpenPGP Key IDs and Fingerprints Registry5.5.4.1. Version 3 Key ID and Fingerprint 🗝️
For a version 3 key, the 8-octet Key ID is made of the low 64 bits of the public modulus of the RSA key.
The fingerprint of a version 3 key is formed by hashing the body (but not the 2-octet length) of the MPIs that form the key material (public modulus n, followed by exponent e) with MD5. Note that both version 3 keys and MD5 are deprecated 🥀.
5.5.4.2. Version 4 Key ID and Fingerprint 🌸
A version 4 fingerprint is the 160-bit SHA-1 hash of the octet 0x99, followed by the 2-octet packet length, followed by the entire Public Key packet starting with the version field. The Key ID is the low-order 64 bits of the fingerprint. Here are the fields of the hash material, including an example of an Ed25519 key 💜:
a.1) 0x99 (1 octet)
a.2) 2-octet, big-endian scalar octet count of (b)-(e)
b) version number = 4 (1 octet)
c) timestamp of key creation (4 octets)
d) algorithm (1 octet): 27 = Ed25519 (example)
e) algorithm-specific fields
Algorithm-specific fields for Ed25519 keys (example):
e.1) 32 octets representing the public key5.5.4.3. Version 6 Key ID and Fingerprint 🌈
A version 6 fingerprint is the 256-bit SHA2-256 hash of the octet 0x9B, followed by the 4-octet packet length, followed by the entire Public Key packet starting with the version field. The Key ID is the high-order 64 bits of the fingerprint. Here are the fields of the hash material, including an example of an Ed25519 key 🧚:
a.1) 0x9B (1 octet)
a.2) 4-octet scalar octet count of (b)-(f)
b) version number = 6 (1 octet)
c) timestamp of key creation (4 octets)
d) algorithm (1 octet): 27 = Ed25519 (example)
e) 4-octet scalar octet count for the key material specified in the
next field
f) algorithm-specific public key material
Algorithm-specific fields for Ed25519 keys (example):
f.1) 32 octets representing the public keyHeads up, collisions can happen 🙈. It's possible for there to be collisions of Key IDs, meaning two different keys with the same Key ID. There is also a much smaller, but still non-zero, probability that two different keys have the same fingerprint.
Also note that if version 3, version 4, and version 6 format keys share the same RSA key material, they will have different Key IDs as well as different fingerprints 🍓.
Finally, the Key ID and fingerprint of a subkey are calculated in the same way as for a primary key, including the 0x99 (version 4 key) or 0x9B (version 6 key) as the first octet. That's true even though this is not a valid Packet Type ID for a public subkey 🤓.
5.5.5. Algorithm-Specific Parts of Keys 🧪
The public and secret key formats specify algorithm-specific parts of a key. The following sections describe them in detail 💖.
5.5.5.1. Algorithm-Specific Part for RSA Keys 🔑
okay so for RSA keys, the public key is just a series of multiprecision integers (MPIs) 💖
- MPI of RSA public modulus n,
- MPI of RSA public encryption exponent e.
The secret key is a series of MPIs too 🎀
- MPI of RSA secret exponent d;
- MPI of RSA secret prime value p;
- MPI of RSA secret prime value q (p < q); and
- MPI of u, the multiplicative inverse of p, mod q.
5.5.5.2. Algorithm-Specific Part for DSA Keys 🌸
for DSA keys, the public key is this series of MPIs ✨
- MPI of DSA prime p;
- MPI of DSA group order q (q is a prime divisor of p-1);
- MPI of DSA group generator g; and
- MPI of DSA public key value y (= g^x mod p where x is secret).
the secret key is just one MPI 🦋
- MPI of DSA secret exponent x.
5.5.5.3. Algorithm-Specific Part for Elgamal Keys 🧚
for Elgamal keys, the public key is this series of MPIs 💜
- MPI of Elgamal prime p;
- MPI of Elgamal group generator g; and
- MPI of Elgamal public key value y (= g^x mod p where x is secret).
and the secret key is a single MPI 🍓
- MPI of Elgamal secret exponent x.
5.5.5.4. Algorithm-Specific Part for ECDSA Keys 💅
for ECDSA keys, the public key is a series of values, bestie 🫶
- A variable-length field containing a curve OID, which is formatted as follows:
- A 1-octet size of the following field; values 0 and 0xFF are reserved for future extensions.
- The octets representing a curve OID, as defined in Section 9.2.
- An MPI of an EC point representing a public key.
The secret key is one single MPI 🔥
- An MPI of an integer representing the secret key, which is a scalar of the public EC point.
5.5.5.5. Algorithm-Specific Part for EdDSALegacy Keys (Deprecated) 🥀
ngl this one is deprecated 😭 but here's what it looks like. For EdDSALegacy keys (deprecated), the public key is this series of values 🌈
- A variable-length field containing a curve OID, formatted as follows:
- A 1-octet size of the following field; values 0 and 0xFF are reserved for future extensions.
- The octets representing a curve OID, as defined in Section 9.2.
- An MPI of an EC point representing a public key Q in prefixed native form (see Section 11.2.2).
The secret key is one single MPI 💜
- An MPI-encoded octet string representing the native form of the secret key in the curve-specific format, as described in Section 9.2.1.
Okay, the important bit ✨ The native form for an EdDSA secret key is a fixed-width sequence of unstructured random octets, with size corresponding to the specific curve. That sequence of random octets is used with a cryptographic digest to produce both a curve-specific secret scalar and a prefix used when making a signature. See Section 5.1.5 of [RFC8032] for more details about how to use the native octet strings for Ed25519Legacy. The value stored in an OpenPGP EdDSALegacy Secret Key packet is the original sequence of random octets.
Also, the only curve defined for use with EdDSALegacy is the Ed25519Legacy OID. That's it, that's the tea 🍵
5.5.5.6. Algorithm-Specific Part for ECDH Keys 🦋
for ECDH keys, the public key is a series of values 🌸
- A variable-length field containing a curve OID, which is formatted as follows:
- A 1-octet size of the following field; values 0 and 0xFF are reserved for future extensions.
- The octets representing a curve OID, as defined in Section 9.2.
- An MPI of an EC point representing a public key, in the point format associated with the curve, as specified in Section 9.2.1.
- A variable-length field containing key derivation function (KDF) parameters, which is formatted as follows:
- A 1-octet size of the following fields; values 0 and 0xFF are reserved for future extensions.
- A 1-octet value 1, reserved for future extensions.
- A 1-octet hash function ID used with a KDF.
- A 1-octet algorithm ID for the symmetric algorithm that is used to wrap the symmetric key for message encryption; see Section 11.5 for details.
The secret key is one single MPI 🔐
- An MPI representing the secret key, in the curve-specific format described in Section 9.2.1.
5.5.5.6.1. ECDH Secret Key Material 🎀
when the curves NIST P-256, NIST P-384, NIST P-521, brainpoolP256r1, brainpoolP384r1, or brainpoolP512r1 are used in ECDH, their secret keys are just a simple integer in standard MPI form 💖 Other curves are presented on the wire differently (though still as a single MPI), as described below and in Section 9.2.1.
5.5.5.6.1.1. Curve25519Legacy ECDH Secret Key Material (Deprecated) 🥺
ok so a Curve25519Legacy secret key is stored as a standard integer in big-endian MPI form. Curve25519Legacy MUST NOT be used in key packets version 6 or above. This is giving reverse octet order compared to the little-endian "native" form found in [RFC7748] 🙈
Also, the integer for a Curve25519Legacy secret key for OpenPGP MUST have the appropriate form; that is, it MUST be divisible by 8, MUST be at least 2^254, and MUST be less than 2^255. The length of this MPI in bits is by definition always 255, so the two leading octets of the MPI will always be 00 FF, and reversing the following 32 octets from the wire will produce the "native" form. 🧠
When generating a new Curve25519Legacy secret key from 32 fully random octets, the following pseudocode produces the MPI wire format (note the similarity to decodeScalar25519 as described in [RFC7748]) 👑
def curve25519Legacy_MPI_from_random(octet_list):
octet_list[0] &= 248
octet_list[31] &= 127
octet_list[31] |= 64
mpi_header = [ 0x00, 0xFF ]
return mpi_header || reversed(octet_list)5.5.5.7. Algorithm-Specific Part for X25519 Keys ⭐
for X25519 keys, the public key is this single value 💜
- 32 octets of the native public key.
The secret key is this single value 🍓
- 32 octets of the native secret key.
See Section 6.1 of [RFC7748] for more details about how to use the native octet strings. The value stored in an OpenPGP X25519 Secret Key packet is the original sequence of random octets. The value stored in an OpenPGP X25519 Public Key packet is the value X25519(secretKey, 9). 🦋
5.5.5.8. Algorithm-Specific Part for X448 Keys 🌷
for X448 keys, the public key is this single value 💅
- 56 octets of the native public key.
The secret key is this single value ✨
- 56 octets of the native secret key.
See Section 6.2 of [RFC7748] for more details about how to use the native octet strings. The value stored in an OpenPGP X448 Secret Key packet is the original sequence of random octets. The value stored in an OpenPGP X448 Public Key packet is the value X448(secretKey, 5). 🫧
5.5.5.9. Algorithm-Specific Part for Ed25519 Keys 🌈
for Ed25519 keys, the public key is this single value 💖
- 32 octets of the native public key.
The secret key is this single value 🔥
- 32 octets of the native secret key.
See Section 5.1.5 of [RFC8032] for more details about how to use the native octet strings. The value stored in an OpenPGP Ed25519 Secret Key packet is the original sequence of random octets. 🧚
5.5.5.10. Algorithm-Specific Part for Ed448 Keys 🌙
for Ed448 keys, the public key is this single value 🍓
- 57 octets of the native public key.
The secret key is this single value 👑
- 57 octets of the native secret key.
See Section 5.2.5 of [RFC8032] for more details about how to use the native octet strings. The value stored in an OpenPGP Ed448 Secret Key packet is the original sequence of random octets. 💜
5.6. Compressed Data Packet (Type ID 8) 🗜️✨
okay the Compressed Data packet contains compressed data 🎁 Typically, this packet is found as the contents of an encrypted packet, or following a Signature or One-Pass Signature packet, and contains a Literal Data packet. It's giving nested-gift energy 🎀
The body of this packet consists of:
- One octet specifying the algorithm used to compress the packet.
- Compressed data, which makes up the remainder of the packet.
A Compressed Data packet's body contains data that is a compression of a series of OpenPGP packets. See Section 10 for details on how messages are formed. 📦
A ZIP-compressed series of packets is compressed into raw DEFLATE blocks [RFC1951]. 🌀
A ZLIB-compressed series of packets is compressed with raw ZLIB-style blocks [RFC1950]. 🌊
A BZip2-compressed series of packets is compressed using the BZip2 [BZ2] algorithm. 🫧
An implementation that generates a Compressed Data packet MUST use the OpenPGP format for packet framing (see Section 4.2.1). It MUST NOT generate a Compressed Data packet with Legacy format (Section 4.2.2). Those are the rules, no cap 🙅♀️💅
An implementation that deals with either historic data or data generated by legacy implementations predating support for [RFC2440] MAY interpret Compressed Data packets that use the Legacy format for packet framing. So that's the flex for the old-school crowd 🕰️💜
5.7. Symmetrically Encrypted Data Packet (Type ID 9) 🔐
okay so this packet holds data that's been encrypted with a symmetric key algorithm 🧚 once you decrypt it, you get other packets inside, usually a Literal Data packet or a compressed data packet. in theory it could be any sequence of packets that forms a valid OpenPGP Message, which is wild 💅
🚨 this packet is obsolete, bestie. An implementation MUST NOT create this packet. An implementation SHOULD reject such a packet and stop processing the message. If an implementation chooses to process the packet anyway, it MUST return a clear warning that a non-integrity-protected packet has been processed.
the packet format can't really be handled safely in general, because the ciphertext it gives you is malleable 💀 see Section 13.7 for a better OpenPGP encryption container that doesn't have this flaw 🦋
the body of this packet has two parts:
- a random prefix. it's block-size random octets (so 16 octets for a 128-bit block length), followed by a copy of the last two octets, all encrypted together in Cipher Feedback (CFB) mode with an IV of all zeros 🌸
- the data, encrypted in CFB mode, using the last block-size octets of the first ciphertext as the IV ✨
the symmetric cipher can be specified in a Public Key or Symmetric Key Encrypted Session Key packet that comes before the Symmetrically Encrypted Data packet. when that happens, the cipher algorithm ID gets prefixed to the session key before it's encrypted 🔑 if no packets of those types come before the encrypted data, the IDEA algorithm is used, with the session key calculated as the MD5 hash of the passphrase. ngl that's deprecated, so please don't do that 🙈
the data is encrypted in CFB mode (see Section 12.9). for the random prefix, the IV is all zeros. instead of getting randomized encryption through an IV, you encrypt a string whose length is the block size of the cipher plus two 🎀 the first block-size octets (for example, 16 octets for a 128-bit block length) are random, and the next two octets are copies of the last two octets of those random octets. for example, with a 16-octet block length, octet 17 is a copy of octet 15, and octet 18 is a copy of octet 16. for a cipher with a block length of 8, octet 9 is a copy of octet 7, and octet 10 is a copy of octet 8. (in both examples we're counting the first octet as number 1 🧮)
after you encrypt those block-size-plus-two octets, you make a new CFB context for the data, with the last block-size octets of the first ciphertext as the IV. (alternatively, and equivalently, the CFB state gets resynchronized: the last block-size octets of ciphertext go through the cipher, and the block boundary resets 🔄)
the repeated two octets in the random prefix let the receiver check right away whether the session key is wrong. this is the "quick check" 🫶 see Section 13.4 for hints on using it properly ✨
5.8. Marker Packet (Type ID 10) 🏷️
the body of the Marker packet is:
- the three octets
0x50,0x47,0x50, which spell "PGP" in UTF-8 🥰
such a packet MUST be ignored when received. no exceptions, it's giving "do not engage" 🙅♀️
5.9. Literal Data Packet (Type ID 11) 📄
a Literal Data packet holds the body of a message, meaning data that's not supposed to be interpreted any further 💖
the body of this packet has four parts:
- a 1-octet field that says how the data is formatted.
- if it's
b(0x62), the Literal Data packet contains binary data 🤖 - if it's
u(0x75), the Literal Data packet contains UTF-8-encoded text, so it might need line ends converted to local form or other text mode changes 📝 - previous versions of the OpenPGP specification used
t(0x74) for textual data, but they didn't specify the character encoding. so implementations SHOULD NOT emit this value 🙃 an implementation that receives a Literal Data packet with this value in the format field SHOULD interpret the packet data as UTF-8 encoded text, unless reliable (not attacker-controlled) context indicates a specific alternate text encoding. this mode is deprecated because it's so ambiguous 😭 - some implementations predating [RFC2440] also defined a value of
las a "local" mode for machine-local conversions. [RFC1991] incorrectly says this local mode flag is 1 (ASCII numeral one). both of these local modes are deprecated 💀
- if it's
- the file name as a string (1-octet length, followed by a file name). it can be a zero-length string. usually, if the source of the encrypted data is a file, this is the name of the encrypted file 📁 an implementation MAY consider the file name in the Literal Data packet to be a more authoritative name than the actual file name 👑
- a 4-octet number that indicates a date associated with the literal data. usually that's the modification date of a file, the time the packet was created, or a zero that means no specific time ⏰
- the rest of the packet is the literal data itself.
text data MUST be encoded with UTF-8 (see [RFC3629]) and stored with <CR><LF> text endings (that's network-normal line endings). the receiving implementation should convert these to native line endings 🌈
now, the tea ☕ OpenPGP signatures do not include the formatting octet, the file name, or the date field of the Literal Data packet in a signature hash. so those fields aren't protected against tampering in a signed document. a receiving implementation MUST NOT treat those fields as though the surrounding signature cryptographically secures them, whether it's showing them to the user or acting on them 🚫
because they're inherently malleable, an implementation that generates a Literal Data packet SHOULD avoid storing any significant data in these fields 🫠 if the implementation is certain the data is textual and encoded with UTF-8 (for example, if it's going to follow this Literal Data packet with a Signature packet of type 0x01 (see Section 5.2.1)), it MAY set the format octet to u. otherwise, it MUST set the format octet to b. it SHOULD set the filename to the empty string (encoded as a single zero octet) and the timestamp to zero (encoded as four zero octets) 🧼
an application that wants filesystem metadata inside a signature is advised to sign an encapsulated archive (for example, [PAX]) 📦
an implementation that generates a Literal Data packet MUST use the OpenPGP format for packet framing (see Section 4.2.1). it MUST NOT generate a Literal Data packet with Legacy format (Section 4.2.2) 🙅♀️
an implementation that deals with either historic data or data generated by an implementation that predates support for [RFC2440] MAY interpret Literal Data packets that use the Legacy format for packet framing 🕰️
5.9.1. Special Filename _CONSOLE (Deprecated) 🙊
the Literal Data packet's filename field has a historical special case for the name _CONSOLE. when the filename field is _CONSOLE, the message counts as "for your eyes only" 👀 this means the message data is unusually sensitive, and the receiving program should process it more carefully, maybe by not saving the received data to disk, for example 🕵️♀️
an OpenPGP deployment that generates Literal Data packets MUST NOT depend on this indicator being honored in any particular way. it can't be enforced, and the field isn't covered by any cryptographic signature, so it's not really a bodyguard 💅
it is NOT RECOMMENDED to use this special filename in a newly generated Literal Data packet 🚫
5.10. Trust Packet (Type ID 12) 🛡️
the Trust packet is only used within keyrings, and it's not normally exported. Trust packets hold data that records the user's choices about which keyholders are trustworthy introducers, plus other information the implementation uses for trust 🤝 the format of Trust packets is defined by each implementation, so it's kinda a "do your own thing" situation 🌸
Trust packets SHOULD NOT be emitted to output streams that get transferred to other users, and they SHOULD be ignored on any input other than local keyring files 🔒
5.11. User ID Packet (Type ID 13) 🪪
a User ID packet is UTF-8 text meant to represent the name and email address of the keyholder 💌 by convention it includes a mail name-addr as described in [RFC2822], but there are no restrictions on its content, so it's very much a free-for-all 🎀 the packet length in the header tells you the length of the User ID 📏
5.12. User Attribute Packet (Type ID 17) 🖼️
the User Attribute packet is a variation of the User ID packet. it can store more kinds of data than the User ID packet, which is limited to text 🌈 like the User ID packet, a User Attribute packet can be certified by the key owner ("self-signed") or by any other key owner who cares to certify it. except where noted, a User Attribute packet can go anywhere a User ID packet can 👑
User Attribute packets aren't a required part of the OpenPGP specification, but implementations SHOULD provide at least enough compatibility to properly handle a certification signature on the User Attribute packet 🫶 a simple way to do this is to treat the User Attribute packet as a User ID packet with opaque contents, but an implementation may use any method it wants ✨
the User Attribute packet is made up of one or more attribute subpackets. each subpacket has a subpacket header and a body. the header consists of:
- the subpacket length (1, 2, or 5 octets)
- the Subpacket Type ID (1 octet)
and then the subpacket-specific data follows 🧁
here's the table of currently known subpackets 💜
+=========+=============================+================+
| ID | Attribute Subpacket | Reference |
+=========+=============================+================+
| 0 | Reserved | |
+---------+-----------------------------+----------------+
| 1 | Image Attribute Subpacket | Section 5.12.1 |
+---------+-----------------------------+----------------+
| 100-110 | Private or Experimental Use | |
+---------+-----------------------------+----------------+
Table 13: OpenPGP User Attribute Subpacket Types Registryan implementation SHOULD ignore any subpacket of a type that it doesn't recognize 🙈 that's the vibe, bestie 🦋
5.12.1. Image Attribute Subpacket ✨
Okay so the Image Attribute subpacket is how you stash an image, presumably (but not required to be) a pic of the key owner 📸🎀
It starts with an image header. The first two octets of that header hold the header's length. ngl, this is a total historical accident: unlike every other multi-octet number in this doc, this one is little-endian 🙈 The length is followed by one octet for the header version. The only version defined so far is 1, which is a 16-octet header, so the first three octets of a version 1 header are 0x10, 0x00, 0x01 💅
+=========+================+
| Version | Reference |
+=========+================+
| 1 | Section 5.12.1 |
+---------+----------------+
Table 14: OpenPGP Image
Attribute Versions
RegistryThe fourth octet of a version 1 header says which encoding the image uses. The only one defined is 1, meaning JPEG. Image format IDs 100 through 110 are reserved for Private or Experimental Use 🧪 The rest of the version 1 header is 12 reserved octets, and they all MUST be set to 0 🫶
+=========+=============================+
| ID | Encoding |
+=========+=============================+
| 0 | Reserved |
+---------+-----------------------------+
| 1 | JPEG [JFIF] |
+---------+-----------------------------+
| 100-110 | Private or Experimental Use |
+---------+-----------------------------+
Table 15: OpenPGP Image Attribute
Encoding Format RegistryEverything after the header is the image itself. JPEG is the only image type defined, so the image is stored in the JPEG File Interchange Format (JFIF), which is a standard file format for JPEG images [JFIF] 🌸
An implementation MAY try to figure out the image type by looking at the image data itself, if it can't handle a particular header version or if the encoding format value isn't recognized. Resourceful, honestly 🔥
5.13. Symmetrically Encrypted and Integrity Protected Data Packet (Type ID 18) 🔐
The SEIPD packet holds integrity-protected, encrypted data. Once it's decrypted, it contains other packets that form an OpenPGP Message (see Section 10.3) 🦋
The first octet is always the version number, and different versions hold ciphertext structured in different ways. Version 1 holds data encrypted with a symmetric key algorithm and is protected against modification by the SHA-1 hash algorithm. This mechanism came in with [RFC4880] and offers some protection against ciphertext malleability 🧁
Version 2 holds data encrypted with an AEAD construction. That's a more cryptographically rigorous defense against ciphertext malleability, so it's giving stronger security energy ✨ See Section 13.7 for more on choosing between these formats.
Any new version of the SEIPD packet should be registered in the registry established in Section 10.3.2.1 📝
5.13.1. Version 1 Symmetrically Encrypted and Integrity Protected Data Packet Format 🎀
A version 1 SEIPD packet has two parts:
- A 1-octet version number with value 1.
- Encrypted data -- the output of the selected symmetric key cipher operating in CFB mode.
The symmetric cipher MUST be specified in a Public Key or Symmetric Key Encrypted Session Key packet that comes before the SEIPD packet. Either way, the cipher algorithm ID gets prefixed to the session key before it's encrypted 🔑
The data is encrypted in CFB mode (see Section 12.9). The IV is specified as all zeros. Instead of getting randomized encryption from an IV, OpenPGP sticks an octet string on the front of the data before encrypting it. The length of that string is the cipher's block size in octets, plus two. The first octets, equal in length to the block size, are random. The last two octets are each copies of their 2nd preceding octet. For example, with a cipher whose block size is 128 bits (16 octets), the prefix holds 16 random octets, then two more octets that copy the 15th and 16th octets, respectively. Unlike the deprecated Symmetrically Encrypted Data packet (Section 5.7), this prefix is encrypted in the same CFB context, and no special CFB resynchronization happens 🌈
That repeated 16 bits in the random prefix lets the receiver immediately check whether the session key is wrong. Quick check, no cap 👀 See Section 13.4 for hints on using this "quick check" properly.
Two constant octets, 0xD3 and 0x14, get appended to the plaintext. Then the plaintext goes through the SHA-1 hash function. The input to the hash is the prefix data described above plus all of the plaintext, including the trailing constant octets 0xD3, 0x14. The 20 octets of the SHA-1 hash are then appended to the plaintext (after the constant octets 0xD3, 0x14) and encrypted along with everything else in the same CFB context. This trailing checksum is called the Modification Detection Code (MDC) 🛡️
When decrypting, the plaintext should be hashed with SHA-1, including the prefix data and the trailing constant octets 0xD3, 0x14, but excluding the last 20 octets, which hold the SHA-1 hash. The computed hash is then compared with those last 20 octets of plaintext. A mismatch means the message has been modified, and it MUST be treated as a security problem. Any failure SHOULD be reported to the user 🚨
NON-NORMATIVE EXPLANATION 💅
The MDC system, which is the integrity protection mechanism of the version 1 SEIPD packet, was built to give you integrity that's weaker than a signature but stronger than bare CFB encryption. It's giving middle ground, bestie 🫶
CFB has a limitation: damage to the ciphertext corrupts the affected cipher blocks and the block after them. Also, if data gets removed from the end of a CFB-encrypted block, that removal is undetectable. (CBC mode has a similar limitation, but data removed from the front of the block is undetectable.) 😬
The obvious fix is to digitally sign the encrypted block. But lots of people don't want to habitually sign their data, for a bunch of reasons that are beyond the scope of this document. Suffice it to say that a lot of people value things like deniability just as much as integrity 🙈
So OpenPGP gets more security than raw encryption while keeping deniability, using the MDC system. An MDC is intentionally not a Message Authentication Code (MAC). The name was not picked by accident. It's analogous to a checksum, and that's the whole point 🎯
It's a pretty modest system, but it has proved itself in the real world. It's a solid defense against several attacks that have surfaced since it was created. It has met its modest goals admirably, iykyk 👑
Because it's a modest security system, it has modest requirements on the hash function(s) it uses. It doesn't rely on the hash being collision-free. It relies on the hash being one-way. Say a forger, Frank, wants to send Alice an unsigned message that says, "I've always secretly loved you, signed Bob." It's way easier for him to construct a brand-new message than to modify anything intercepted from Bob 💌 (Note also that if Bob wants to talk to Alice in secret, without authentication or identification, and with a threat model that includes forgers, he has a problem that goes beyond cryptography. Not me, Bob 💀)
Also, unlike nearly every other OpenPGP subsystem, the MDC system has no parameters. It hard-defines SHA-1 as its hash function. That's not an accident. It's a deliberate choice to avoid downgrade and cross-grade attacks while keeping the system simple and fast. (A downgrade attack would replace SHA2-256 with SHA-1, for example. A cross-grade attack would replace SHA-1 with another 160-bit hash, such as RIPEMD-160, for example.) 🧠
However, no update will be needed, because the MDC has been replaced by the AEAD encryption described in this document 🦋✨
5.13.2. Version 2 Symmetrically Encrypted and Integrity Protected Data Packet Format ✨
Okay bestie, the v2 SEIPD packet is giving main character energy 💖 A v2 packet is made up of:
- A 1-octet version number with value 2.
- A 1-octet cipher algorithm ID.
- A 1-octet AEAD algorithm identifier.
- A 1-octet chunk size.
- 32 octets of salt. The salt is used to derive the message key and MUST be securely generated (see Section 13.10). 🧂
- Encrypted data, which is the output of the selected symmetric key cipher running in the given AEAD mode.
- A final summary authentication tag for the AEAD mode.
The decrypted session key and the salt get used to derive an M-bit message key plus N-64 bits for the IV, where M is the key size of the symmetric algorithm and N is the nonce size of the AEAD algorithm. So M + N - 64 bits get derived with HKDF (see [RFC5869]) 🔑 The leftmost M bits become the symmetric algorithm key, and the leftover N - 64 bits become the IV. HKDF is used with SHA256 [RFC6234] as the hash algorithm. The session key is the IKM and the salt is the salt. The Packet Type ID in OpenPGP format encoding (bits 7 and 6 are set, and bits 5-0 carry the Packet Type ID), the version number, the cipher algorithm ID, the AEAD algorithm ID, and the chunk size octet all go in as the info parameter. It's so detailed, I'm obsessed 😭
The KDF mechanism gives you key separation between cipher and AEAD algorithms. Also, an implementation can securely reply to a message even when a recipient's certificate is unknown, by reusing the Encrypted Session Key packets and replying with a different salt that yields a new, unique message key 🦋 See Section 13.8 for guidance on how applications can securely implement this feature.
A v2 SEIPD packet consists of one or more chunks of data. The plaintext of each chunk is of a size specified by the chunk size octet, using the method specified below ✨
The encrypted data is the encryption of each chunk of plaintext, followed immediately by the relevant authentication tag. If the last chunk of plaintext is smaller than the chunk size, the ciphertext for that data may be shorter. Even so, it is followed by a full authentication tag 💅
For each chunk, the AEAD construction is given the Packet Type ID encoded in OpenPGP format (bits 7 and 6 are set, and bits 5-0 carry the Packet Type ID), the version number, the cipher algorithm ID, the AEAD algorithm ID, and the chunk size octet as additional data. For example, the additional data of the first chunk using EAX and AES-128 with a chunk size of 2^22 octets is the octets 0xD2, 0x02, 0x07, 0x01, and 0x10 🎀
After the final chunk, the AEAD algorithm is used to produce a final authentication tag that encrypts the empty string. This AEAD instance is given the additional data specified above, plus an 8-octet, big-endian value specifying the total number of plaintext octets encrypted. This allows detection of a truncated ciphertext 🙈
The chunk size octet specifies the size of chunks using the following formula (in C [C99]), where c is the chunk size octet:
chunk_size = (uint32_t) 1 << (c + 6)An implementation MUST accept chunk size octets with values from 0 to 16. An implementation MUST NOT create data with a chunk size octet value larger than 16 (4 MiB chunks). 🚫
The nonce for AEAD mode consists of two parts. Let N be the size of the nonce. The leftmost N - 64 bits are the IV derived using HKDF. The rightmost 64 bits are the chunk index as a big-endian value. The index of the first chunk is zero 🧚
5.13.3. EAX Mode 🌸
The EAX AEAD algorithm used in this document is defined in [EAX].
The EAX algorithm can only use block ciphers with 16-octet blocks. The nonce is 16 octets long. EAX authentication tags are 16 octets long.
5.13.4. OCB Mode 🍓
The OCB AEAD algorithm used in this document is defined in [RFC7253].
The OCB algorithm can only use block ciphers with 16-octet blocks. The nonce is 15 octets long. OCB authentication tags are 16 octets long.
5.13.5. GCM Mode 🌈
The GCM AEAD algorithm used in this document is defined in [SP800-38D].
The GCM algorithm can only use block ciphers with 16-octet blocks. The nonce is 12 octets long. GCM authentication tags are 16 octets long.
5.14. Padding Packet (Type ID 21) 🫧
The Padding packet contains random data and can be used to defend against traffic analysis (see Section 13.11) on v2 SEIPD messages (see Section 5.13.2) and Transferable Public Keys (see Section 10.1). It's a bit of a sneaky privacy move 🕵️♀️
Such a packet MUST be ignored when received.
Its contents SHOULD be random octets, to make the length obfuscation it provides more robust even when compressed. 🎲
An implementation adding padding to an OpenPGP stream SHOULD place such a packet:
- At the end of a version 6 Transferable Public Key that is transferred over an encrypted channel (see Section 10.1).
- As the last packet of an Optionally Padded Message within a version 2 Symmetrically Encrypted and Integrity Protected Data packet (see Section 10.3.1).
An implementation MUST be able to process Padding packets anywhere else in an OpenPGP stream, so that future revisions of this document may specify further locations for padding. Future-proof era 🔮
Policy about how large to make such a packet to defend against traffic analysis is beyond the scope of this document. Not my problem, but I respect the honesty 💀
6. Base64 Conversions 🎀
As stated in the introduction, OpenPGP's underlying representation for objects is a stream of arbitrary octets, and some systems want these objects to be immune to damage caused by character set translation, data conversions, and so on 🫶
In principle, any printable encoding scheme that met the requirements of the unsafe channel would suffice, since it would not change the underlying binary bit streams of the OpenPGP data structures. The OpenPGP specification defines one such printable encoding scheme to ensure interoperability; see Section 6.2 ✨
The encoding is made of two parts: a base64 encoding of the binary data and an optional checksum. The base64 encoding used is described in Section 4 of [RFC4648], and it is wrapped into lines of no more than 76 characters each.
When decoding base64, an OpenPGP implementation MUST ignore all whitespace.
6.1. Optional Checksum 🧁
The optional checksum is a 24-bit Cyclic Redundancy Check (CRC) converted to four characters of base64 encoding by the same MIME base64 transformation, preceded by an equal sign (=). The CRC is computed using the generator 0x864CFB and an initialization of 0xB704CE. The accumulation is done on the data before it is converted to base64, rather than on the converted data. A sample implementation of this algorithm is in Section 6.1.1 🔢
If present, the checksum with its leading equal sign MUST appear on the next line after the base64-encoded data.
An implementation MUST NOT reject an OpenPGP object when the CRC24 footer is present, missing, malformed, or disagrees with the computed CRC24 sum. When forming ASCII Armor, the CRC24 footer SHOULD NOT be generated, unless interoperability with implementations that require the CRC24 footer to be present is a concern. 🙅♀️
The CRC24 footer MUST NOT be generated if the context or the OpenPGP object being encoded can determine that the consuming implementation accepts base64-encoded blocks without a CRC24 footer. Notably:
- An ASCII-armored Encrypted Message packet sequence that ends in a v2 SEIPD packet MUST NOT contain a CRC24 footer.
- An ASCII-armored sequence of Signature packets that only includes version 6 Signature packets MUST NOT contain a CRC24 footer.
- An ASCII-armored Transferable Public Key packet sequence of a version 6 key MUST NOT contain a CRC24 footer.
- An ASCII-armored keyring consisting of only version 6 keys MUST NOT contain a CRC24 footer.
Rationale: Previous draft versions of this document stated that the CRC24 footer is optional, but the text was ambiguous. In practice, very few implementations require the CRC24 footer to be present. Computing the CRC24 incurs a significant cost while providing no meaningful integrity protection, so generating it is now discouraged. Honestly, it's giving "why were we doing this" 😭🫠
6.1.1. An Implementation of the CRC24 in "C" 💻
The following code is written in [C99].
#define CRC24_INIT 0xB704CEL
#define CRC24_GENERATOR 0x864CFBL
typedef unsigned long crc24;
crc24 crc_octets(unsigned char *octets, size_t len)
{
crc24 crc = CRC24_INIT;
int i;
while (len--) {
crc ^= (*octets++) << 16;
for (i = 0; i < 8; i++) {
crc <<= 1;
if (crc & 0x1000000) {
crc &= 0XFFFFFF; /* Clear bit 25 to avoid overflow */
crc ^= CRC24_GENERATOR;
}
}
}
return crc & 0xFFFFFFL;
}6.2. Forming ASCII Armor 🦋
When OpenPGP encodes data into ASCII Armor, it puts specific headers around the base64-encoded data, so OpenPGP can reconstruct the data later. An OpenPGP implementation MAY use ASCII Armor to protect raw binary data. OpenPGP tells the user what kind of data is encoded in the ASCII Armor through the use of the headers 📜
Concatenating the following data creates ASCII Armor:
- An Armor Header Line, appropriate for the type of data
- Armor Headers
- A blank (zero length or containing only whitespace) line
- The ASCII-Armored data
- An optional Armor Checksum (discouraged; see Section 6.1)
- The Armor Tail, which depends on the Armor Header Line
6.2.1. Armor Header Line 🏷️
An Armor Header Line consists of the appropriate header line text surrounded by five (5) dashes (-, 0x2D) on either side of the header line text. The header line text is chosen based on the type of data being encoded in Armor and how it is being encoded. Header line texts include the following strings:
+===================+============================================+
| Armor Header | Use |
+===================+============================================+
| BEGIN PGP MESSAGE | Used for signed, encrypted, or compressed |
| | files. |
+-------------------+--------------------------------------------+
| BEGIN PGP PUBLIC | Used for armoring public keys. |
| KEY BLOCK | |
+-------------------+--------------------------------------------+
| BEGIN PGP PRIVATE | Used for armoring private keys. |
| KEY BLOCK | |
+-------------------+--------------------------------------------+
| BEGIN PGP | Used for detached signatures, OpenPGP/MIME |
| SIGNATURE | signatures, and cleartext signatures. |
+-------------------+--------------------------------------------+
Table 16: OpenPGP Armor Header Lines RegistryNote that all of these Armor Header Lines are to consist of a complete line. Therefore, the header lines MUST start at the beginning of a line, and MUST NOT have text other than whitespace following them on the same line. Lines up, no extra stuff, periodt 💅✨
6.2.2. Armor Headers ✨
Okay bestie, Armor Headers are pairs of strings that can give the user or the receiving OpenPGP implementation some useful info about how to decode or use the message 💅 Here's the tea: they're part of the armor, not the message, so they're not protected by any signatures applied to the message 🙈
Each Armor Header is a key-value pair. A colon (: 0x3A) and a single space (0x20) separate the key from the value 🎀 An OpenPGP implementation may consider improperly formatted Armor Headers to be a corruption of the ASCII Armor, but it SHOULD make an effort to recover 🫶 Unknown keys should be silently ignored, and an implementation SHOULD continue to process the message 🌸
Note that some transport methods are sensitive to line length. For example, the SMTP protocol that carries email has a line length limit of 998 characters (see Section 2.1.1 of [RFC5322]) 📬
While the base64 data has a limit of 76 characters (Section 6), there is no limit on the length of Armor Headers 🦋 Care should be taken to make sure the Armor Headers are short enough to survive transport. One way to do this is to repeat an Armor Header Key multiple times, each with a different value, so that no single line gets overly long 🧚
Currently defined Armor Header Keys are these:
+=========+==============================+=================+
| Key | Summary | Reference |
+=========+==============================+=================+
| Version | Implementation information | Section 6.2.2.1 |
+---------+------------------------------+-----------------+
| Comment | Arbitrary text | Section 6.2.2.2 |
+---------+------------------------------+-----------------+
| Hash | Hash algorithms used in some | Section 6.2.2.3 |
| | v4 cleartext signed messages | |
+---------+------------------------------+-----------------+
| Charset | Character set | Section 6.2.2.4 |
+---------+------------------------------+-----------------+
Table 17: OpenPGP Armor Header Keys Registry6.2.2.1. "Version" Armor Header
The Armor Header Key Version describes the OpenPGP implementation and version used to encode the message. To minimize metadata, implementations SHOULD NOT emit this key and its corresponding value except for debugging purposes with explicit user consent 🔍🫣
6.2.2.2. "Comment" Armor Header
The Armor Header Key Comment supplies a user-defined comment 💬 OpenPGP defines all text to be in UTF-8, so a comment may be any UTF-8 string. BUT the whole point of armoring is to provide 7-bit clean data, so if a comment has characters outside the ASCII range of UTF-8, they may very well not survive transport 😭
6.2.2.3. "Hash" Armor Header
The Armor Header Key Hash is deprecated, but some older implementations still expect it in messages using the Cleartext Signature Framework (Section 7) 🕰️ When it's present, this key contains a comma-separated list of hash algorithms used in the signatures on the message, with digest names as specified in the "Text Name" column in Table 23. These headers SHOULD NOT be emitted unless:
- the cleartext signed message contains a version 4 signature made using a SHA2-based digest (SHA224, SHA256, SHA384, or SHA512), and
- the cleartext signed message might be verified by a legacy OpenPGP implementation that requires this header.
A verifying application MUST decline to validate any signature in a message with a non-conformant Hash header (that is, a Hash header that contains anything other than a comma-separated list of hash algorithms) 🚫 When a conformant Hash header is present, a verifying application MUST ignore its contents and defer to the hash algorithm indicated in the Embedded Signature 🎯
6.2.2.4. "Charset" Armor Header
The Armor Header Key Charset describes the character set that the plaintext is in (see [RFC2978]) 🌍 Please note that OpenPGP defines text to be in UTF-8, and an implementation will get the best results by translating into and out of UTF-8. However, that's easier said than done in lots of places 😤 There are also communities of users who have no need for UTF-8 because they're all happy with a character set like ISO Latin-5 or a Japanese character set. In those cases, an implementation MAY override the UTF-8 default by using this header key. An implementation MAY implement this key and any translations it cares to, and an implementation MAY ignore it and assume all text is UTF-8 🌸
6.2.3. Armor Tail Line
The Armor Tail Line is composed the same way as the Armor Header Line, except the string "BEGIN" is replaced by the string "END" 🏁
7. Cleartext Signature Framework 🖊️
Sometimes you want to sign a textual octet stream without ASCII armoring the stream itself, so the signed text is still readable with any tool that can render text 👀 To bind a signature to such a cleartext, the Cleartext Signature Framework is used, and it follows the same basic format and restrictions as the ASCII armoring described in Section 6.2. (Note that this framework is not intended to be reversible. [RFC3156] defines another way to sign cleartext messages for environments that support MIME.)
7.1. Cleartext Signed Message Structure
An OpenPGP cleartext signed message consists of:
- The cleartext header
-----BEGIN PGP SIGNED MESSAGE-----on a single line. - One or more legacy Hash Armor Headers that MAY be included by some implementations and MUST be ignored when well formed (see Section 6.2.2.3).
- An empty line (not included in the message digest).
- The dash-escaped cleartext.
- A line ending separating the cleartext and following armored signature (not included in the message digest).
- The ASCII-armored signature(s), including the
-----BEGIN PGP SIGNATURE-----Armor Header and Armor Tail Lines.
As with any other Text signature (Section 5.2.1.2), a cleartext signature is calculated on the text using canonical <CR><LF> line endings 📏 As described above, the line ending before the -----BEGIN PGP SIGNATURE----- Armor Header Line of the armored signature is not considered part of the signed text 🧐
Also, any trailing whitespace (spaces 0x20 and tabs 0x09) at the end of any line is removed before signing or verifying a cleartext signed message ✂️
Between the -----BEGIN PGP SIGNED MESSAGE----- line and the first empty line, the only Armor Header permitted is a well-formed Hash Armor Header (see Section 6.2.2.3). To reduce the risk of confusion about what has been signed, a verifying implementation MUST decline to validate any signature in a cleartext message if that message has any other Armor Header present in this location 🛑
7.2. Dash-Escaped Text
The cleartext content of the message must also be dash-escaped 🙅♀️
Dash-escaped cleartext is ordinary cleartext where every line starting with a "-" (HYPHEN-MINUS, U+002D) is prefixed by the sequence "-" (HYPHEN-MINUS, U+002D) and " " (SPACE, U+0020). This stops the parser from mistaking the cleartext itself for Armor Headers 🎭 An implementation MAY dash-escape any line, SHOULD dash-escape lines commencing in "From" followed by a space, and MUST dash-escape any line commencing in a dash. The message digest is computed using the cleartext itself, not the dash-escaped form 🧮
When reversing dash-escaping, an implementation MUST strip the string - if it occurs at the beginning of a line, and it SHOULD provide a warning for - and any character other than a space at the beginning of a line ⚠️
7.3. Issues with the Cleartext Signature Framework
Creating a cleartext signed message involves trimming trailing whitespace on every line, so the Cleartext Signature Framework will fail to safely round-trip any textual stream that may include semantically meaningful whitespace 😩
For example, the Unified Diff format [UNIFIED-DIFF] contains semantically meaningful whitespace: an empty line of context will consist of a line with a single " " (SPACE, U+0020) character, and any line that has trailing whitespace added or removed will represent such a change with semantically meaningful whitespace 🧩
Also, a very large Cleartext Signature Framework message is unlikely to work well 🐘 It'll be hard for any human reading it to know which part is covered by the signature, because they can't take in the whole message at once, especially if an Armor Header line sits somewhere in the body. And very large messages can't be processed in a single pass, since the signature salt and digest algorithms are only discovered at the end 🙃
An implementation that knows it's working with a textual stream with any of the above characteristics SHOULD NOT use the Cleartext Signature Framework. Safe alternatives for a semantically meaningful OpenPGP signature over such a file format are:
- A signed message, as described in Section 10.3.
- A detached signature, as described in Section 10.4.
Either of these alternatives may be ASCII-armored (see Section 6.2) if they need to be sent across a text-only (or 7-bit clean) channel 📡
Finally, when a Cleartext Signature Framework message is presented to the user as is, an attacker can include additional text in the Hash header, which may mislead the user into thinking it's part of the signed text 👀🚨 The signature validation constraints described in Sections 6.2.2.3 and 7.1 help to reduce the risk of arbitrary or misleading text in the Armor Headers 🛡️
8. Regular Expressions 🔤
This section describes Regular Expressions 🧠
- Regular Expression: Zero or more branches, separated by
|. It matches anything that matches one of the branches. - Branch: Zero or more pieces, concatenated. It matches a match for the first, followed by a match for the second, and so on.
- Piece: An atom possibly followed by
*,+, or?. An atom followed by*matches a sequence of 0 or more matches of the atom. An atom followed by+matches a sequence of 1 or more matches of the atom. An atom followed by?matches a match of the atom or the null string. - Atom: A Regular Expression in parentheses (matching a match for the Regular Expression), a range (see below), a
.(matching any single Unicode character), a^(matching the null string at the beginning of the input string), a$(matching the null string at the end of the input string), a\followed by a single Unicode character (matching that character), or a single Unicode character with no other significance (matching that character). - Range: A sequence of characters enclosed in
[]. It normally matches any single character from the sequence. If the sequence begins with^, it matches any single Unicode character not from the rest of the sequence. If two characters in the sequence are separated by-, this is shorthand for the full list of Unicode characters between them in codepoint order (for example,[0-9]matches any decimal digit). To include a literal]in the sequence, make it the first character (following a possible^). To include a literal-, make it the first or last character.
9. Constants ✨
Okay so this section is the constants lineup 🎀 aka all the ID numbers and names OpenPGP uses for its algorithms, and it's giving organized.
Quick heads up: these tables are not exhaustive lists 👀. An implementation MAY implement an algorithm that isn't on these lists, as long as its algorithm ID comes from the Private or Experimental Use algorithm range 🧚.
See Section 12 for more discussion of the algorithms 💜.
9.1. Public Key Algorithms 🔑
+===+==============+=========+============+===========+=============+
| ID| Algorithm |Public | Secret Key | Signature | PKESK |
| | |Key | Format | Format | Format |
| | |Format | | | |
+===+==============+=========+============+===========+=============+
| 0| Reserved | | | | |
+---+--------------+---------+------------+-----------+-------------+
| 1| RSA (Encrypt |MPI(n), | MPI(d), | MPI(m^d | MPI(m^e |
| | or Sign) |MPI(e) | MPI(p), | mod n) | mod n) |
| | [FIPS186] |[Section | MPI(q), | [Section | [Section |
| | |5.5.5.1] | MPI(u) | 5.2.3.1] | 5.1.3] |
+---+--------------+---------+------------+-----------+-------------+
| 2| RSA Encrypt- |MPI(n), | MPI(d), | N/A | MPI(m^e |
| | Only |MPI(e) | MPI(p), | | mod n) |
| | [FIPS186] |[Section | MPI(q), | | [Section |
| | |5.5.5.1] | MPI(u) | | 5.1.3] |
+---+--------------+---------+------------+-----------+-------------+
| 3| RSA Sign- |MPI(n), | MPI(d), | MPI(m^d | N/A |
| | Only |MPI(e) | MPI(p), | mod n) | |
| | [FIPS186] |[Section | MPI(q), | [Section | |
| | |5.5.5.1] | MPI(u) | 5.2.3.1] | |
+---+--------------+---------+------------+-----------+-------------+
| 16| Elgamal |MPI(p), | MPI(x) | N/A | MPI(g^k |
| | (Encrypt- |MPI(g), | | | mod p), |
| | Only) |MPI(y) | | | MPI(m * |
| | [ELGAMAL] |[Section | | | y^k mod |
| | |5.5.5.3] | | | p) |
| | | | | | [Section |
| | | | | | 5.1.4] |
+---+--------------+---------+------------+-----------+-------------+
| 17| DSA (Digital |MPI(p), | MPI(x) | MPI(r), | N/A |
| | Signature |MPI(q), | | MPI(s) | |
| | Algorithm) |MPI(g), | | [Section | |
| | [FIPS186] |MPI(y) | | 5.2.3.2] | |
| | |[Section | | | |
| | |5.5.5.2] | | | |
+---+--------------+---------+------------+-----------+-------------+
| 18| ECDH public |OID, | MPI(value | N/A | MPI(point |
| | key |MPI(point| in curve- | | in curve- |
| | algorithm |in curve-| specific | | specific |
| | |specific | format) | | point |
| | |format), | 9.2.1] | | format), |
| | |KDFParams| | | size |
| | |[Sections| | | octet, |
| | |9.2.1 and| | | encoded |
| | |5.5.5.6] | | | key |
| | | | | | [Sections |
| | | | | | 9.2.1, |
| | | | | | 5.1.5, |
| | | | | | and 11.5] |
+---+--------------+---------+------------+-----------+-------------+
| 19| ECDSA public |OID, | MPI(value) | MPI(r), | N/A |
| | key |MPI(point| | MPI(s) | |
| | algorithm |in SEC1 | | [Section | |
| | [FIPS186] |format) | | 5.2.3.2] | |
| | |[Section | | | |
| | |5.5.5.4] | | | |
+---+--------------+---------+------------+-----------+-------------+
| 20| Reserved | | | | |
| | (formerly | | | | |
| | Elgamal | | | | |
| | Encrypt or | | | | |
| | Sign) | | | | |
+---+--------------+---------+------------+-----------+-------------+
| 21| Reserved for | | | | |
| | Diffie- | | | | |
| | Hellman | | | | |
| | (X9.42, as | | | | |
| | defined for | | | | |
| | IETF-S/MIME) | | | | |
+---+--------------+---------+------------+-----------+-------------+
| 22| EdDSALegacy |OID, | MPI(value | MPI, MPI | N/A |
| | (deprecated) |MPI(point| in curve- | [Sections | |
| | |in | specific | 9.2.1 and | |
| | |prefixed | format) | 5.2.3.3] | |
| | |native | [Section | | |
| | |format) | 9.2.1] | | |
| | |[Sections| | | |
| | |11.2.2 | | | |
| | |and | | | |
| | |5.5.5.5] | | | |
+---+--------------+---------+------------+-----------+-------------+
| 23| Reserved | | | | |
| | (AEDH) | | | | |
+---+--------------+---------+------------+-----------+-------------+
| 24| Reserved | | | | |
| | (AEDSA) | | | | |
+---+--------------+---------+------------+-----------+-------------+
| 25| X25519 |32 octets| 32 octets | N/A | 32 |
| | |[Section | | | octets, |
| | |5.5.5.7] | | | size |
| | | | | | octet, |
| | | | | | encoded |
| | | | | | key |
| | | | | | [Section |
| | | | | | 5.1.6] |
+---+--------------+---------+------------+-----------+-------------+
| 26| X448 |56 octets| 56 octets | N/A | 56 |
| | |[Section | | | octets, |
| | |5.5.5.8] | | | size |
| | | | | | octet, |
| | | | | | encoded |
| | | | | | key |
| | | | | | [Section |
| | | | | | 5.1.7] |
+---+--------------+---------+------------+-----------+-------------+
| 27| Ed25519 |32 octets| 32 octets | 64 octets | |
| | |[Section | | [Section | |
| | |5.5.5.9] | | 5.2.3.4] | |
+---+--------------+---------+------------+-----------+-------------+
| 28| Ed448 |57 octets| 57 octets | 114 | |
| | |[Section | | octets | |
| | |5.5.5.10]| | [Section | |
| | | | | 5.2.3.5] | |
+---+--------------+---------+------------+-----------+-------------+
|100| Private or | | | | |
| to| Experimental | | | | |
|110| Use | | | | |
+---+--------------+---------+------------+-----------+-------------+ Table 18: OpenPGP Public Key Algorithms RegistryBestie, the must-haves 🫶: implementations MUST implement Ed25519 (27) for signatures and X25519 (25) for encryption. Implementations SHOULD implement Ed448 (28) and X448 (26) too, so that's extra credit 🌈.
RSA (1) keys are deprecated 🥀 and SHOULD NOT be generated, but they may be interpreted. RSA Encrypt-Only (2) and RSA Sign-Only (3) are deprecated and MUST NOT be generated (see Section 12.4). Elgamal (16) keys are deprecated and MUST NOT be generated (see Section 12.6). DSA (17) keys are deprecated and MUST NOT be generated (see Section 12.5). For notes on Elgamal Encrypt or Sign (20) and X9.42 (21), see Section 12.8. Implementations MAY implement any other algorithm 💅.
History corner 📜: an implementation conforming to the previous version of this specification [RFC4880] has only DSA (17) and Elgamal (16) as the algorithms that MUST be implemented.
A compatible specification of ECDSA is given in [RFC6090] (as "KT-I Signatures") and in [SEC1]. ECDH is defined in Section 11.5 of this document 🦋.
9.2. ECC Curves for OpenPGP ✨
Okay so, this is the part where we meet the curves! 🦋 The parameter curve OID is an array of octets that defines a named curve. Literally the ID card for each curve, bestie. 💅
The table below specifies the exact sequence of octets for each named curve referenced in this document. It also specifies which public key algorithms the curve can be used with, as well as the size of expected elements in octets. Note that there is a break in "EdDSALegacy" for display purposes only. 🎀
+======================+===+========+================+======+=======+
|ASN.1 Object |OID| Curve |Curve Name |Usage |Field |
|Identifier |Len| OID | | |Size |
| | | Octets | | |(fsize)|
+======================+===+========+================+======+=======+
|1.2.840.10045.3.1.7 |8 | 2A 86 |NIST P-256 |ECDSA,|32 |
| | | 48 CE | |ECDH | |
| | | 3D 03 | | | |
| | | 01 07 | | | |
+----------------------+---+--------+----------------+------+-------+
|1.3.132.0.34 |5 | 2B 81 |NIST P-384 |ECDSA,|48 |
| | | 04 00 | |ECDH | |
| | | 22 | | | |
+----------------------+---+--------+----------------+------+-------+
|1.3.132.0.35 |5 | 2B 81 |NIST P-521 |ECDSA,|66 |
| | | 04 00 | |ECDH | |
| | | 23 | | | |
+----------------------+---+--------+----------------+------+-------+
|1.3.36.3.3.2.8.1.1.7 |9 | 2B 24 |brainpoolP256r1 |ECDSA,|32 |
| | | 03 03 | |ECDH | |
| | | 02 08 | | | |
| | | 01 01 | | | |
| | | 07 | | | |
+----------------------+---+--------+----------------+------+-------+
|1.3.36.3.3.2.8.1.1.11 |9 | 2B 24 |brainpoolP384r1 |ECDSA,|48 |
| | | 03 03 | |ECDH | |
| | | 02 08 | | | |
| | | 01 01 | | | |
| | | 0B | | | |
+----------------------+---+--------+----------------+------+-------+
|1.3.36.3.3.2.8.1.1.13 |9 | 2B 24 |brainpoolP512r1 |ECDSA,|64 |
| | | 03 03 | |ECDH | |
| | | 02 08 | | | |
| | | 01 01 | | | |
| | | 0D | | | |
+----------------------+---+--------+----------------+------+-------+
|1.3.6.1.4.1.11591.15.1|9 | 2B 06 |Ed25519Legacy |EdDSA |32 |
| | | 01 04 | |Legacy| |
| | | 01 DA | | | |
| | | 47 0F | | | |
| | | 01 | | | |
+----------------------+---+--------+----------------+------+-------+
|1.3.6.1.4.1.3029.1.5.1|10 | 2B 06 |Curve25519Legacy|ECDH |32 |
| | | 01 04 | | | |
| | | 01 97 | | | |
| | | 55 01 | | | |
| | | 05 01 | | | |
+----------------------+---+--------+----------------+------+-------+
Table 19: OpenPGP ECC Curve OIDs and Usage RegistryThe "Field Size (fsize)" column represents the field size of the group in number of octets, rounded up, such that x or y coordinates for a point on the curve or native point representations for the curve can be represented in that many octets. The curves specified here, and scalars such as the base point order and the private key, can be represented in fsize octets. However, note that curves exist outside this specification where the representation of scalars requires an additional octet. 🧚
The sequence of octets in the third column is the result of applying the Distinguished Encoding Rules (DER) to the ASN.1 Object Identifier with subsequent truncation. The truncation removes the two fields of encoded Object Identifier. The first omitted field is 1 octet representing the Object Identifier tag, and the second omitted field is the length of the Object Identifier body. For example, the complete ASN.1 DER encoding for the NIST P-256 curve OID is "06 08 2A 86 48 CE 3D 03 01 07", from which the first entry in the table above is constructed by omitting the first two octets. Only the truncated sequence of octets is the valid representation of a curve OID. 🤓
The deprecated OIDs for Ed25519Legacy and Curve25519Legacy are used only in version 4 keys and signatures. Implementations MAY implement these variants for compatibility with existing version 4 key material and signatures. Implementations MUST NOT accept or generate version 6 key material using the deprecated OIDs. 🚫
9.2.1. Curve-Specific Wire Formats 🌸
Some elliptic curve public key algorithms use different conventions for specific fields depending on the curve in use. Each field is always formatted as an MPI, but with a curve-specific framing. This table summarizes those distinctions. 💜
+================+========+============+=======+=========+==========+
|Curve |ECDH |ECDH Secret |EdDSA |EdDSA |EdDSA |
| |Point |Key MPI |Secret |Signature|Signature |
| |Format | |Key MPI|first MPI|second |
| | | | | |MPI |
+================+========+============+=======+=========+==========+
|NIST P-256 |SEC1 |integer |N/A |N/A |N/A |
+----------------+--------+------------+-------+---------+----------+
|NIST P-384 |SEC1 |integer |N/A |N/A |N/A |
+----------------+--------+------------+-------+---------+----------+
|NIST P-521 |SEC1 |integer |N/A |N/A |N/A |
+----------------+--------+------------+-------+---------+----------+
|brainpoolP256r1 |SEC1 |integer |N/A |N/A |N/A |
+----------------+--------+------------+-------+---------+----------+
|brainpoolP384r1 |SEC1 |integer |N/A |N/A |N/A |
+----------------+--------+------------+-------+---------+----------+
|brainpoolP512r1 |SEC1 |integer |N/A |N/A |N/A |
+----------------+--------+------------+-------+---------+----------+
|Ed25519Legacy |N/A |N/A |32 |32 octets|32 octets |
| | | |octets |of R |of S |
| | | |of | | |
| | | |secret | | |
+----------------+--------+------------+-------+---------+----------+
|Curve25519Legacy|prefixed|integer |N/A |N/A |N/A |
| |native |(Section | | | |
| | |5.5.5.6.1.1)| | | |
+----------------+--------+------------+-------+---------+----------+
Table 20: OpenPGP ECC Curve-Specific Wire Formats RegistryFor the native octet-string forms of Ed25519Legacy values, see [RFC8032]. For the native octet-string forms of Curve25519Legacy secret scalars and points, see [RFC7748]. 📚
9.3. Symmetric Key Algorithms 🔐
+=========+============================================+
| ID | Algorithm |
+=========+============================================+
| 0 | Plaintext or unencrypted data |
+---------+--------------------------------------------+
| 1 | IDEA [IDEA] |
+---------+--------------------------------------------+
| 2 | TripleDES (or DES-EDE) [SP800-67] with |
| | 168-bit key derived from 192 |
+---------+--------------------------------------------+
| 3 | CAST5 with 128-bit key [RFC2144] |
+---------+--------------------------------------------+
| 4 | Blowfish with 128-bit key, 16 rounds |
| | [BLOWFISH] |
+---------+--------------------------------------------+
| 5 | Reserved |
+---------+--------------------------------------------+
| 6 | Reserved |
+---------+--------------------------------------------+
| 7 | AES with 128-bit key [AES] |
+---------+--------------------------------------------+
| 8 | AES with 192-bit key |
+---------+--------------------------------------------+
| 9 | AES with 256-bit key |
+---------+--------------------------------------------+
| 10 | Twofish with 256-bit key [TWOFISH] |
+---------+--------------------------------------------+
| 11 | Camellia with 128-bit key [RFC3713] |
+---------+--------------------------------------------+
| 12 | Camellia with 192-bit key |
+---------+--------------------------------------------+
| 13 | Camellia with 256-bit key |
+---------+--------------------------------------------+
| 100-110 | Private or Experimental Use |
+---------+--------------------------------------------+
| 253-255 | Reserved to avoid collision with Secret |
| | Key Encryption (Table 2 and Section 5.5.3) |
+---------+--------------------------------------------+
Table 21: OpenPGP Symmetric Key Algorithms RegistryOkay this one is so important, bestie! 🔥 Implementations MUST implement AES-128. Implementations SHOULD implement AES-256. Implementations MUST NOT encrypt data with IDEA, TripleDES, or CAST5. 🙅♀️ Implementations MAY decrypt data that uses IDEA, TripleDES, or CAST5 for the sake of reading older messages or new messages from implementations predating support for [RFC2440]. An Implementation that decrypts data using IDEA, TripleDES, or CAST5 SHOULD generate a deprecation warning about the symmetric algorithm, indicating that message confidentiality is suspect. Implementations MAY implement any other algorithm. 🌈
9.4. Compression Algorithms 🗜️✨
okay so compression algorithms, bestie, let's get into it 💅 this table is the registry of IDs and what each one means, and it's giving very organized energy 🎀
+=========+=============================+
| ID | Algorithm |
+=========+=============================+
| 0 | Uncompressed |
+---------+-----------------------------+
| 1 | ZIP [RFC1951] |
+---------+-----------------------------+
| 2 | ZLIB [RFC1950] |
+---------+-----------------------------+
| 3 | BZip2 [BZ2] |
+---------+-----------------------------+
| 100-110 | Private or Experimental Use |
+---------+-----------------------------+
Table 22: OpenPGP Compression
Algorithms RegistrySo the rules, no cap: implementations MUST implement uncompressed data 🫶 Implementations SHOULD implement ZLIB 🌸 For interoperability reasons, implementations SHOULD be able to decompress using ZIP 💖 And implementations MAY implement any other algorithm 🦋
9.5. Hash Algorithms 🔐💜
okay this one is a whole vibe 😭 the hash algorithm registry, with the text names and the V6 signature salt sizes all lined up 🎀
+=========+==================+=============+========================+
| ID | Algorithm | Text Name | V6 Signature |
| | | | Salt Size |
+=========+==================+=============+========================+
| 0 | Reserved | | |
+---------+------------------+-------------+------------------------+
| 1 | MD5 [RFC1321] | "MD5" | N/A |
+---------+------------------+-------------+------------------------+
| 2 | SHA-1 [FIPS180] | "SHA1" | N/A |
+---------+------------------+-------------+------------------------+
| 3 | RIPEMD-160 | "RIPEMD160" | N/A |
| | [RIPEMD-160] | | |
+---------+------------------+-------------+------------------------+
| 4 | Reserved | | |
+---------+------------------+-------------+------------------------+
| 5 | Reserved | | |
+---------+------------------+-------------+------------------------+
| 6 | Reserved | | |
+---------+------------------+-------------+------------------------+
| 7 | Reserved | | |
+---------+------------------+-------------+------------------------+
| 8 | SHA2-256 | "SHA256" | 16 |
| | [FIPS180] | | |
+---------+------------------+-------------+------------------------+
| 9 | SHA2-384 | "SHA384" | 24 |
| | [FIPS180] | | |
+---------+------------------+-------------+------------------------+
| 10 | SHA2-512 | "SHA512" | 32 |
| | [FIPS180] | | |
+---------+------------------+-------------+------------------------+
| 11 | SHA2-224 | "SHA224" | 16 |
| | [FIPS180] | | |
+---------+------------------+-------------+------------------------+
| 12 | SHA3-256 | "SHA3-256" | 16 |
| | [FIPS202] | | |
+---------+------------------+-------------+------------------------+
| 13 | Reserved | | |
+---------+------------------+-------------+------------------------+
| 14 | SHA3-512 | "SHA3-512" | 32 |
| | [FIPS202] | | |
+---------+------------------+-------------+------------------------+
| 100-110 | Private or | | |
| | Experimental Use | | |
+---------+------------------+-------------+------------------------+
Table 23: OpenPGP Hash Algorithms Registryand then there's the second table, which is the OID and full hash prefix for each hash, for RSA signatures using EMSA-PKCS1-v1_5 padding 🌈 it's a lot of hex, but it's the kind of detail that makes it iconic 🙈
+============+=========================+=========================+
| Hash | OID | Full Hash Prefix |
| Algorithm | | |
+============+=========================+=========================+
| MD5 | 1.2.840.113549.2.5 | 0x30, 0x20, 0x30, 0x0C, |
| | | 0x06, 0x08, 0x2A, 0x86, |
| | | 0x48, 0x86, 0xF7, 0x0D, |
| | | 0x02, 0x05, 0x05, 0x00, |
| | | 0x04, 0x10 |
+------------+-------------------------+-------------------------+
| SHA-1 | 1.3.14.3.2.26 | 0x30, 0x21, 0x30, 0x09, |
| | | 0x06, 0x05, 0x2B, 0x0E, |
| | | 0x03, 0x02, 0x1A, 0x05, |
| | | 0x00, 0x04, 0x14 |
+------------+-------------------------+-------------------------+
| RIPEMD-160 | 1.3.36.3.2.1 | 0x30, 0x21, 0x30, 0x09, |
| | | 0x06, 0x05, 0x2B, 0x24, |
| | | 0x03, 0x02, 0x01, 0x05, |
| | | 0x00, 0x04, 0x14 |
+------------+-------------------------+-------------------------+
| SHA2-256 | 2.16.840.1.101.3.4.2.1 | 0x30, 0x31, 0x30, 0x0D, |
| | | 0x06, 0x09, 0x60, 0x86, |
| | | 0x48, 0x01, 0x65, 0x03, |
| | | 0x04, 0x02, 0x01, 0x05, |
| | | 0x00, 0x04, 0x20 |
+------------+-------------------------+-------------------------+
| SHA2-384 | 2.16.840.1.101.3.4.2.2 | 0x30, 0x41, 0x30, 0x0D, |
| | | 0x06, 0x09, 0x60, 0x86, |
| | | 0x48, 0x01, 0x65, 0x03, |
| | | 0x04, 0x02, 0x02, 0x05, |
| | | 0x00, 0x04, 0x30 |
+------------+-------------------------+-------------------------+
| SHA2-512 | 2.16.840.1.101.3.4.2.3 | 0x30, 0x51, 0x30, 0x0D, |
| | | 0x06, 0x09, 0x60, 0x86, |
| | | 0x48, 0x01, 0x65, 0x03, |
| | | 0x04, 0x02, 0x03, 0x05, |
| | | 0x00, 0x04, 0x40 |
+------------+-------------------------+-------------------------+
| SHA2-224 | 2.16.840.1.101.3.4.2.4 | 0x30, 0x2D, 0x30, 0x0D, |
| | | 0x06, 0x09, 0x60, 0x86, |
| | | 0x48, 0x01, 0x65, 0x03, |
| | | 0x04, 0x02, 0x04, 0x05, |
| | | 0x00, 0x04, 0x1C |
+------------+-------------------------+-------------------------+
| SHA3-256 | 2.16.840.1.101.3.4.2.8 | 0x30, 0x31, 0x30, 0x0D, |
| | | 0x06, 0x09, 0x60, 0x86, |
| | | 0x48, 0x01, 0x65, 0x03, |
| | | 0x04, 0x02, 0x08, 0x05, |
| | | 0x00, 0x04, 0x20 |
+------------+-------------------------+-------------------------+
| SHA3-512 | 2.16.840.1.101.3.4.2.10 | 0x30, 0x51, 0x30, 0x0D, |
| | | 0x06, 0x09, 0x60, 0x86, |
| | | 0x48, 0x01, 0x65, 0x03, |
| | | 0x04, 0x02, 0x0a, 0x05, |
| | | 0x00, 0x04, 0x40 |
+------------+-------------------------+-------------------------+
Table 24: OpenPGP Hash Algorithm Identifiers for RSA
Signatures' Use of EMSA-PKCS1-v1_5 Padding RegistryNow the rules, and these ones are SERIOUS bestie 💀 Implementations MUST implement SHA2-256 🔥 Implementations SHOULD implement SHA2-384 and SHA2-512 ✨ Implementations MAY implement other algorithms 🧚 Implementations SHOULD NOT create messages that require the use of SHA-1, with the exception of computing version 4 key fingerprints for purposes of the MDC in version 1 Symmetrically Encrypted and Integrity Protected Data packets 🙈
Implementations MUST NOT generate signatures with MD5, SHA-1, or RIPEMD-160 💀 Implementations MUST NOT use MD5, SHA-1, or RIPEMD-160 as a hash function in an ECDH KDF 💀 Implementations MUST NOT generate packets using MD5, SHA-1, or RIPEMD-160 as a hash function in an S2K KDF 💀 Implementations MUST NOT decrypt a secret using MD5, SHA-1, or RIPEMD-160 as a hash function in an S2K KDF in a version 6 (or later) packet 💀 Implementations MUST NOT validate any recent signature that depends on MD5, SHA-1, or RIPEMD-160 💀
And the last one is the nuance one 🥺 Implementations SHOULD NOT validate any old signature that depends on MD5, SHA-1, or RIPEMD-160 unless the signature's creation date predates known weakness of the algorithm used, and the implementation is confident that the message has been in the secure custody of the user the whole time 🦋
9.6. AEAD Algorithms 🎀💖
last registry, and it's the AEAD one, bestie ✨ it lists the IDs, the names, and the nonce and authentication tag lengths in octets 🌸
+=========+==================+==============+====================+
| ID | Name | Nonce Length | Authentication Tag |
| | | (Octets) | Length (Octets) |
+=========+==================+==============+====================+
| 0 | Reserved | | |
+---------+------------------+--------------+--------------------+
| 1 | EAX [EAX] | 16 | 16 |
+---------+------------------+--------------+--------------------+
| 2 | OCB [RFC7253] | 15 | 16 |
+---------+------------------+--------------+--------------------+
| 3 | GCM [SP800-38D] | 12 | 16 |
+---------+------------------+--------------+--------------------+
| 100-110 | Private or | | |
| | Experimental Use | | |
+---------+------------------+--------------+--------------------+
Table 25: OpenPGP AEAD Algorithms RegistryAnd the rule for this one is simple 🫶 Implementations MUST implement OCB 💜 Implementations MAY implement EAX, GCM, and other algorithms 🌈
10. Packet Sequence Composition ✨
okay so OpenPGP packets don't just get thrown together 🙅♀️ they get assembled into sequences to make messages and transfer keys. not every packet order is valid or makes sense, so this section lays out the rules for how packets should be placed into sequences 💅
There are three distinct sequences of packets:
- Transferable Public Keys (Section 10.1) and their close counterpart, Transferable Secret Keys (Section 10.2) 💖
- OpenPGP Messages (Section 10.3) 📩
- Detached Signatures (Section 10.4) ✍️
Each sequence has an explicit grammar of which packet types (Table 3) can show up where. if there's an unknown critical packet, or a known packet that just isn't expected there, that's a critical error and it kills the whole sequence (see Section 4.3) 💀. unknown non-critical packets, though, can appear anywhere in any sequence 🦋 that's the structured way new packets get introduced while certain packets stay strictly handled 👑
An implementation may "recognize" a packet without actually implementing it. Packet Criticality is how the producer tells the consumer whether a brand-new unknown packet should throw an error or just get ignored 🙈
Note that previous versions of this document didn't have Packet Criticality and didn't give clear guidance on unknown packets. so implementations of those older versions may reject unknown non-critical packets or accept unknown critical packets 🫠
When generating a sequence according to one of the three grammars, an implementation MUST NOT inject a critical packet of a type that doesn't follow the grammar 🚫
When consuming a sequence, if an implementation hits a critical packet of an inappropriate type for the relevant grammar, it MUST reject the sequence with an error. 🙅♀️ bye
10.1. Transferable Public Keys
okay, public keys get shared, obviously 🔑 this section explains the structure of public keys in transit so everyone's on the same page 🫶 A Transferable Public Key is also known as an OpenPGP certificate. that name tells it apart from its Public Key packets (Sections 5.5.1.1 and 5.5.1.2) and the underlying cryptographic key material 🧁
10.1.1. OpenPGP Version 6 Certificate Structure
here's the layout for a version 6 certificate 🎀 square brackets mean optional, and ellipses mean repeats:
Primary Key
[Revocation Signature...]
Direct Key Signature...
[User ID or User Attribute
[Certification Revocation Signature...]
[Certification Signature...]]...
[Subkey [Subkey Revocation Signature...]
Subkey Binding Signature...]...
[Padding]on top of all that, a Marker packet (Section 5.8) can appear anywhere in the sequence 🌸
Note that a version 6 key uses a self-signed Direct Key signature to store algorithm preferences ✨
Every subkey for a version 6 primary key MUST be a version 6 subkey. Every subkey MUST have at least one Subkey Binding signature. Every Subkey Binding signature MUST be a self-signature (made by the version 6 primary key, obvs). And like all the others, every self-signature made by a version 6 key MUST be a version 6 signature 😭
10.1.2. OpenPGP Version 6 Revocation Certificate
when a primary version 6 Public Key gets revoked, it's sometimes distributed with only the Revocation Signature:
Primary Key
Revocation Signaturein this case the Direct Key signature isn't needed anymore, since the primary key itself is already marked as unusable 🪦
10.1.3. OpenPGP Version 4 Certificate Structure
here's the version 4 layout 🍓
Primary Key
[Revocation Signature]
[Direct Key Signature...]
[User ID or User Attribute [Signature...]]...
[Subkey [Subkey Revocation Signature...]
Subkey Binding Signature...]...and yep, a Marker packet (Section 5.8) can appear anywhere in the sequence here too 🦋
A subkey always has at least one Subkey Binding signature after it, issued using the primary key to tie the two keys together. These binding signatures may be version 3 or version 4, but they SHOULD be version 4 format. Subkeys that can issue signatures MUST have a version 4 binding signature because of the REQUIRED embedded Primary Key Binding signature 💜
Every subkey for a version 4 primary key MUST be a version 4 subkey.
when a primary version 4 Public Key is revoked, the Revocation Signature sometimes gets distributed all by itself, without the primary key packet it applies to. that's called a "revocation certificate". Instead, a version 6 revocation certificate MUST include the primary key packet, as described in Section 10.1.2 🫣
10.1.4. OpenPGP Version 3 Key Structure
version 3 keys are the old-school ones 🕰️ here's the shape:
RSA Public Key
[Revocation Signature]
User ID [Signature...]
[User ID [Signature...]]...a Marker packet (Section 5.8) can also appear anywhere in the sequence 🧚
Each signature certifies the RSA public key and the User ID right before it. The RSA public key can have lots of User IDs, and each User ID can have lots of signatures. Version 3 keys are deprecated, so implementations MUST NOT generate new version 3 keys, but they MAY continue to use existing ones 🥀
Version 3 keys MUST NOT have subkeys.
10.1.5. Common Requirements
the Public Key packet always goes first 🥇
The primary key MUST be an algorithm that's capable of making signatures (so, not an encryption-only algorithm). That's because the primary key needs to be able to create self-signatures (see Section 5.2.3.10). Subkeys, on the other hand, can be keys of any type. for example, you could have a single-key RSA key, an Ed25519 primary key with an RSA encryption subkey, an Ed25519 primary key with an X25519 subkey, and so on 🌈
Each User ID packet provides the identity of the owner of the public key. If there are multiple User ID packets, that's multiple ways of identifying the same unique individual. like, someone might have two email addresses and make a User ID for each one 📧 A Transferable Public Key SHOULD include at least one User ID packet, unless storage requirements prohibit it.
Immediately after each User ID packet, there are zero or more Signature packets. Each Signature packet is calculated over the immediately preceding User ID packet and the initial Public Key packet. The signature certifies the public key and the User ID. Basically, the signer is saying "yeah, I believe this public key belongs to the person this User ID identifies" 🤝
Within the same section as the User ID packets, there are zero or more User Attribute packets. Like User ID packets, a User Attribute packet is followed by zero or more Signature packets, calculated on the immediately preceding User Attribute packet and the initial Public Key packet 🎟️
User Attribute packets and User ID packets can be freely intermixed in this section, as long as the signatures that follow them stay attached to the proper User Attribute or User ID packet 🧩
After the User ID and Attribute packets and their associated signatures, zero or more Subkey packets follow, each with their own signatures. Generally, subkeys show up when the top-level public key is a certification-only key. However, any version 4 or version 6 key may have subkeys, and those subkeys may be encryption keys, signing keys, authentication keys, etc. It's good practice to use separate subkeys for every operation (signature-only, encryption-only, authentication-only, and so on) 💅
Each Subkey packet MUST be followed by one Signature packet, which should be a Subkey Binding signature issued by the top-level key. For subkeys that can issue signatures, the Subkey Binding signature MUST contain an Embedded Signature subpacket with a Primary Key Binding signature (Type ID 0x19) issued by the subkey on the top-level key 🔗
Subkey and Key packets may each be followed by a Revocation Signature packet to indicate that the key is revoked. Revocation Signatures are only accepted if they're issued by the key itself, or by a key that's authorized to issue revocations via a Revocation Key subpacket in a self-signature by the top-level key 🚨
The optional trailing Padding packet is a way to defend against traffic analysis (see Section 13.11) 🕵️♀️ For maximum interoperability, if the Public Key packet is a version 4 key, the optional Padding packet SHOULD NOT be present unless the recipient has indicated that they can ignore it successfully. An implementation that can receive a Transferable Public Key with a version 6 Public Key primary key MUST be able to accept (and ignore) the trailing optional Padding packet 🫧
Transferable Public Key packet sequences may be concatenated, so you can transfer multiple public keys in one operation (see Section 3.6) 📦
10.2. Transferable Secret Keys
okay now the secret stuff 🤫 OpenPGP users may transfer secret keys too. The format of a Transferable Secret Key is the same as a Transferable Public Key, except that Secret Key and Secret Subkey packets can be used in addition to Public Key and Public Subkey packets. If a single Secret Key or Secret Subkey packet is included in a packet sequence, the whole thing is a Transferable Secret Key and should be handled and marked as one (see Section 6.2.1) 🔐
An implementation SHOULD include self-signatures on any User IDs and subkeys, because that lets a complete public key be automatically extracted from the Transferable Secret Key 🪄 An implementation MAY choose to omit the self-signatures, especially if a Transferable Public Key comes along with the Transferable Secret Key 🙌
10.3. OpenPGP Messages
an OpenPGP Message is a packet, or a sequence of packets, that follows these grammar rules 📜 (a comma (,) means sequential composition, and a vertical bar (|) separates alternatives):
OpenPGP Message: Encrypted Message | Signed Message | Compressed
Message | Literal Message.
Compressed Message: Compressed Data Packet.
Literal Message: Literal Data Packet.
ESK: Public Key Encrypted Session Key Packet | Symmetric Key
Encrypted Session Key Packet.
ESK Sequence: ESK | ESK Sequence, ESK.
Encrypted Data: Symmetrically Encrypted Data Packet | Symmetrically
Encrypted and Integrity Protected Data Packet.
Encrypted Message: Encrypted Data | ESK Sequence, Encrypted Data.
One-Pass Signed Message: One-Pass Signature Packet, OpenPGP Message,
Corresponding Signature Packet.
Signed Message: Signature Packet, OpenPGP Message | One-Pass Signed
Message.
Optionally Padded Message: OpenPGP Message | OpenPGP Message,
Padding Packet.and one more time for the people in the back: a Marker packet (Section 5.8) can appear anywhere in the sequence 🎀✨
10.3.1. Unwrapping Encrypted and Compressed Messages ✨
Okay so, on top of the grammar we just went through, some messages can get "unwrapped" to reveal brand new messages inside 🎁 total plot twist. The main ones:
- Decrypting a version 2 Symmetrically Encrypted and Integrity Protected Data packet MUST yield a valid Optionally Padded Message. 💖
- Decrypting a version 1 Symmetrically Encrypted and Integrity Protected Data packet or -- for historic data -- a Symmetrically Encrypted Data packet MUST yield a valid OpenPGP Message. 🦋
- Decompressing a Compressed Data packet MUST also yield a valid OpenPGP Message. 🌸
When any unwrapping happens, the resulting stream of octets gets parsed into a series of OpenPGP packets, just like any other stream of octets 🧚. The packet boundaries in that series are expected to line up with the length of the unwrapped octet stream. Here's the big rule: an implementation MUST NOT interpret octets beyond the boundaries of the unwrapped octet stream as part of any OpenPGP packet. 🙅 And if an implementation hits a packet whose header length says it would stretch past the end of the unwrapped octet stream, the implementation MUST reject that packet as malformed and unusable. No drama, just a hard no 💀
10.3.2. Additional Constraints on Packet Sequences 🎀
Some subtle combinations are technically allowed by this grammar but are still unacceptable 🙈 the grammar says yes, the vibes say no.
10.3.2.1. Packet Versions in Encrypted Messages 🔐
Quick recap: an Encrypted Message is zero or more PKESK packets (Section 5.1) and SKESK packets (Section 5.3), followed by an SEIPD (Section 5.13) payload. In some historic data, the payload may be a deprecated SED packet (Section 5.7) instead of SEIPD, but implementations MUST NOT generate SED packets (see Section 13.7). The versions of the ESK packets before the payload MUST align with the version of the payload SEIPD packet, as described in this section.
v3 PKESK and v4 SKESK packets both carry the Symmetric Cipher Algorithm ID and the session key for the next SEIPD packet in cleartext. A v1 SEIPD doesn't have a symmetric algorithm ID, so all ESK packets before a v1 SEIPD payload MUST be either v3 PKESK or v4 SKESK. 🙃
Now the plot thickens: the cleartext of v6 ESK packets (PKESK or SKESK) doesn't contain a Symmetric Cipher Algorithm ID, so they can't be used with a v1 SEIPD payload. The payload after any v6 PKESK or v6 SKESK packet MUST be a v2 SEIPD. 🧩
Also, to avoid potentially conflicting cipher algorithm IDs (and to keep things simple), implementations MUST NOT precede a v2 SEIPD payload with either v3 PKESK or v4 SKESK packets. 🚫
Here's the cheat sheet for which packet versions show up in an Encrypted Message. An implementation MUST only generate an Encrypted Message using packet versions that match a row with "Yes" in the "Generate?" column. The other rows are only there for historic interoperability. A conforming implementation MUST only generate an Encrypted Message using packets whose versions correspond to a single row. 📋
+==============+=====================+==================+===========+
| Version of | Version of | Version of | Generate? |
| Encrypted | Preceding Symmetric | Preceding | |
| Data Payload | Key ESK (If Any) | Public Key | |
| | | ESK (If Any) | |
+==============+=====================+==================+===========+
| SED (Section | - | v2 PKESK | No |
| 5.7) | | [RFC2440] | |
+--------------+---------------------+------------------+-----------+
| SED (Section | v4 SKESK | v3 PKESK | No |
| 5.7) | (Section 5.3.1) | (Section | |
| | | 5.1.1) | |
+--------------+---------------------+------------------+-----------+
| v1 SEIPD | v4 SKESK | v3 PKESK | Yes |
| (Section | (Section 5.3.1) | (Section | |
| 5.13.1) | | 5.1.1) | |
+--------------+---------------------+------------------+-----------+
| v2 SEIPD | v6 SKESK | v6 PKESK | Yes |
| (Section | (Section 5.3.2) | (Section | |
| 5.13.2) | | 5.1.2) | |
+--------------+---------------------+------------------+-----------+
Table 26: OpenPGP Encrypted Message Packet Versions RegistryAn implementation processing an Encrypted Message MUST discard any preceding ESK packet with a version that does not align with the version of the payload. 🗑️
10.3.2.2. Packet Versions in Signatures ✍️
OpenPGP Key packets and Signature packets are versioned too 🔖. A Signature's version typically matches the version of the signing key. When a version 6 key produces a Signature packet, it MUST produce a version 6 Signature packet, no matter the Signature packet type. When a message is signed or verified using the one-pass construction, the version of the One-Pass Signature packet (Section 5.4) should also be aligned to the other versions.
Some legacy implementations produced unaligned signature versions for older key material. Those are listed in the table below, again just for historic interoperability 🕰️. A conforming implementation MUST only generate Signature packets with version numbers matching rows with "Yes" in the "Generate?" column.
+=====================+================+============+===========+
| Signing Key Version | Signature | OPS Packet | Generate? |
| | Packet Version | Version | |
+=====================+================+============+===========+
| 3 (Section 5.5.2.1) | 3 (Section | 3 (Section | No |
| | 5.2.2) | 5.4) | |
+---------------------+----------------+------------+-----------+
| 4 (Section 5.5.2.2) | 3 (Section | 3 (Section | No |
| | 5.2.2) | 5.4) | |
+---------------------+----------------+------------+-----------+
| 4 (Section 5.5.2.2) | 4 (Section | 3 (Section | Yes |
| | 5.2.3) | 5.4) | |
+---------------------+----------------+------------+-----------+
| 6 (Section 5.5.2.3) | 6 (Section | 6 (Section | Yes |
| | 5.2.3) | 5.4) | |
+---------------------+----------------+------------+-----------+
Table 27: OpenPGP Key and Signature Versions RegistryNote, however, that a version mismatch between these packets doesn't invalidate the packet sequence as a whole. It only invalidates the signature itself 🫠, since a signature with an unknown version SHOULD be discarded (see Section 5.2.5).
10.4. Detached Signatures 📎
Some OpenPGP applications use so-called "detached signatures". For example, a program bundle might contain a file, and alongside it a second file that is a detached signature of the first one 🗂️. Detached signatures are simply one or more Signature packets stored separately from the data they're a signature for.
In addition, a Marker packet (Section 5.8) and a Padding packet (Section 5.14) can show up anywhere in the sequence 🌟.
11. Elliptic Curve Cryptography 🌊
This section describes the algorithms and parameters used with Elliptic Curve Cryptography (ECC) keys. A thorough intro to ECC can be found in [KOBLITZ]. For the methods to generate a uniformly distributed ECC private key, refer to [FIPS186], Appendix B.4 🔑.
None of the ECC methods described in this document are allowed with deprecated version 3 keys. 🙅♀️
11.1. ECC Curves 🌈
This document references three named prime field curves defined in [FIPS186] as "Curve P-256", "Curve P-384", and "Curve P-521", and three named prime field curves defined in [RFC5639] as "brainpoolP256r1", "brainpoolP384r1", and "brainpoolP512r1". All six curves can be used with the ECDSA and ECDH public key algorithms. They're referenced using a sequence of octets, called the curve OID. Section 9.2 describes in detail how that sequence of octets is formed 🧾.
Separate algorithms exist for X25519 and X448 [RFC7748], and for Ed25519 and Ed448 [RFC8032]. On top of that, legacy OIDs are defined for "Curve25519Legacy" (for encryption using the ECDH algorithm) and "Ed25519Legacy" (for signing using the EdDSALegacy algorithm) 🕹️.
11.2. EC Point Wire Formats 🪞
A point on an elliptic curve will always be represented on the wire as an MPI. Each curve uses a specific point format for the data within the MPI itself. Each format uses a designated prefix octet to make sure the high octet has at least 1 bit set, so the MPI stays a constant size 📏.
+=================+================+================+
| Name | Wire Format | Reference |
+=================+================+================+
| SEC1 | 0x04 || x || y | Section 11.2.1 |
+-----------------+----------------+----------------+
| Prefixed native | 0x40 || native | Section 11.2.2 |
+-----------------+----------------+----------------+
Table 28: OpenPGP Elliptic Curve Point Wire
Formats Registry11.2.1. SEC1 EC Point Wire Format
For a SEC1-encoded (uncompressed) point, the content of the MPI is:
B = 04 || x || yHere x and y are the coordinates of the point P = (x, y). Each one is encoded in big-endian format and zero-padded to the adjusted underlying field size. The adjusted underlying field size is the underlying field size rounded up to the nearest 8-bit boundary, as noted in the "fsize" column in Section 9.2. This encoding is compatible with the definition given in [SEC1] ✨.
11.2.2. Prefixed Native EC Point Wire Format
For a custom compressed point, the content of the MPI is:
B = 40 || pHere p is the public key of the point, encoded using the rules defined for the specified curve. This format is used for ECDH keys based on curves expressed in Montgomery form, and for points when using EdDSA 🎀.
11.2.3. Notes on EC Point Wire Formats
Given all of the above, the exact size of the MPI payload for an encoded point is 515 bits for both NIST P-256 and brainpoolP256r1, 771 for both NIST P-384 and brainpoolP384r1, 1059 for NIST P-521, 1027 for brainpoolP512r1, and 263 for both Curve25519Legacy and Ed25519Legacy. For example, the length of an EdDSALegacy public key for the curve Ed25519Legacy is 263 bits: 7 bits to represent the 0x40 prefix octet and 32 octets for the native value of the public key 🧮.
Even though the zero point (also called the "point at infinity") may occur as a result of arithmetic operations on points of an elliptic curve, it SHALL NOT appear in data structures defined in this document 🚫.
Each particular curve uses a designated wire format for the point found in its public key or ECDH data structure. An implementation MUST NOT use a different wire format for a point other than the wire format associated with the curve. Stick to the format for your curve, bestie 💅
11.3. EC Scalar Wire Formats ✨
okay so not every ECC value is a point on a curve 🌸. Things like secret keys and signature components are just scalars, but they still get sent over the wire in OpenPGP as an MPI 🎀.
the tricky part is that different deployments do this differently 💅. Some curves treat these values as opaque bit strings with the high bit set, and others treat them as real integers in the normal OpenPGP big-endian form. which way you go depends on the public key algorithm you're using, so it's not one-size-fits-all 🦋
+==========+===========================================+===========+
| Type | Description | Reference |
+==========+===========================================+===========+
| integer | An integer encoded in big-endian format | Section |
| | as a standard OpenPGP MPI | 3.2 |
+----------+-------------------------------------------+-----------+
| octet | An octet string of fixed length that may | Section |
| string | be shorter on the wire due to leading | 11.3.1 |
| | zeros being stripped by the MPI encoding | |
| | and may need to be zero-padded before use | |
+----------+-------------------------------------------+-----------+
| prefixed | An octet string of fixed length N, | Section |
| N octets | prefixed with octet 0x40 to ensure no | 11.3.2 |
| | leading zero octet | |
+----------+-------------------------------------------+-----------+Table 29: OpenPGP Elliptic Curve Scalar Encodings Registry 🧚
11.3.1. EC Octet String Wire Format
some opaque octet strings go on the wire as an MPI by stripping the leading zeros and just counting the bits that are left 🙈 these strings have a known, fixed length. in this doc we write them as MPI(N octets of X), where N is the length in octets you're expecting.
example time 🍓 a 5-octet opaque string (MPI(5 octets of X)) where X is 00 02 EE 19 00 ends up on the wire as an MPI like this: 00 1A 02 EE 19 00.
to encode X onto the wire, set the MPI's 2-octet bit counter to the highest set bit (bit 26, or 0x001A), and don't send that leading all-zero octet 💖
to undo it, an implementation that knows X should be 5 octets can do this, if it wants:
- make sure the MPI's 2-octet bit count is less than or equal to 40 (5 octets of 8 bits) ✨
- allocate 5 octets and set them all to zero to start 🌈
- copy the MPI data octets (without the two count octets) into the lower octets of that space 🫶
11.3.2. EC Prefixed Octet String Wire Format
another way to keep a fixed-length byte string simple on the wire while still being an MPI is to put a dedicated non-zero octet in front of it 👑 this spec uses 0x40 as that prefix octet. we write it as MPI(prefixed N octets of X), where N is the known length of the byte string 💜
for example, a 5-octet opaque string using MPI(prefixed 5 octets of X) where X is 00 02 EE 19 00 goes onto the wire as: 00 2F 40 00 02 EE 19 00 🦋
to encode it, put the octet 0x40 in front (its 7th bit is set), then set the MPI's 2-octet bit counter to 47 (0x002F, which is 7 bits for the prefix octet plus 40 bits for the string) 🎀
to decode it off the wire, an implementation that knows the variable was built this way can:
- make sure the first three octets of the MPI (the 2-bit count octets plus the prefix octet) are
00 2F 40, and - use the rest of the MPI straight off the wire ✨
note that this is a similar trick to the one used in the EC point encodings in Section 11.2.2 🌸
11.4. Key Derivation Function
okay this one's important 🔥 a key derivation function (KDF) is necessary to implement EC encryption. the Concatenation Key Derivation Function (Approved Alternative 1) [SP800-56A], with the KDF hash function being SHA2-256 [FIPS180] or something stronger, is REQUIRED 💅
for convenience, here's a synopsis of the method. it's got some big simplifications because this doc only allows a restricted set of hash functions 🧚 but [SP800-56A] is the real normative source, so that's the one to trust 🫶
// Implements KDF( X, oBits, Param );
// Input: point X = (x,y)
// oBits - the desired size of output
// hBits - the size of output of hash function Hash
// Param - octets representing the parameters
// Assumes that oBits <= hBits
// Convert the point X to the octet string:
// ZB' = 04 || x || y
// and extract the x portion from ZB'
ZB = x;
MB = Hash ( 00 || 00 || 00 || 01 || ZB || Param );
return oBits leftmost bits of MB.note that ZB in the KDF description above is the compact representation of X as defined in Section 4.2 of [RFC6090] 🌈
11.5. ECDH Algorithm 💖
okay bestie, this section is all about the One-Pass Diffie-Hellman method 🦋. It's two jobs stitched together: the ECC Diffie-Hellman part that makes a shared secret, and the key derivation part that turns that secret into a derived key ✨. Then it adds a key wrapping method that uses the derived key to protect a session key, which is the thing that actually encrypts your message 🔐. it's giving teamwork 🫶
The One-Pass Diffie-Hellman method C(1, 1, ECC CDH) [SP800-56A] MUST be implemented with the following restrictions: the ECC Cofactor Diffie-Hellman (CDH) primitive used by this method is modified to always assume the cofactor is 1, the KDF specified in Section 11.4 is used, and the KDF parameters specified below are used 💅
The KDF parameters are one concatenation of 5 fields, some variable-length and some fixed-length. They're built to match the OtherInfo bit string definition in [SP800-56A] 🌸
- A variable-length field containing a curve OID, formatted like this:
- A 1-octet size of the following field.
- The octets representing a curve OID, as defined in Section 9.2.
- A 1-octet public key algorithm ID, as defined in Section 9.1.
- A variable-length field containing KDF parameters. These are identical to the matching field in the ECDH public key, and they're formatted like this:
- A 1-octet size of the following fields; values 0 and 0xFF are reserved for future extensions.
- A 1-octet value 0x01, reserved for future extensions.
- A 1-octet hash function ID used with the KDF.
- A 1-octet algorithm ID for the symmetric algorithm that is used to wrap the symmetric key for message encryption; see Section 11.5 for details.
- 20 octets representing the UTF-8 encoding of the string "Anonymous Sender" padded at the end with spaces (0x20) to 20 octets. That's the octet sequence
41 6E 6F 6E 79 6D 6F 75 73 20 53 65 6E 64 65 72 20 20 20 20🧚 - A variable-length field containing the fingerprint of the recipient encryption subkey, which identifies the key material needed for decryption. For version 4 keys this field is 20 octets. For version 6 keys it's 32 octets.
The size in octets of the KDF parameters sequence, as defined above, depends on the key and curve. For encrypting to a version 4 key it's 54 for curve NIST P-256, 51 for curves NIST P-384 and NIST P-521, 55 for curves brainpoolP256r1, brainpoolP384r1, and brainpoolP512r1, or 56 for Curve25519Legacy. For encrypting to a version 6 key, the sequence is 66 for curve NIST P-256, 63 for curves NIST P-384 and NIST P-521, or 67 for curves brainpoolP256r1, brainpoolP384r1, and brainpoolP512r1 🔥
The key wrapping method is described in [RFC3394]. The KDF makes a symmetric key, and that key is used as a KEK as specified in [RFC3394]. Check out Section 11.5.1 for how to pick the KEK algorithm, which SHOULD be one of the three AES algorithms. Key wrapping and unwrapping both use the default initial value of [RFC3394] ✨
To make the input for the key wrapping method, you first concatenate these values:
- The 1-octet algorithm identifier, if it was passed (in the case of a v3 PKESK packet).
- The session key.
- A 2-octet checksum of the session key, equal to the sum of the session key octets, modulo 65536.
After that, the above values get padded to an 8-octet granularity using the method described in [RFC8018] 🎀
For example, in a version 3 Public Key Encrypted Session Key packet, an AES-256 session key is encoded like this, forming a 40-octet sequence:
09 k0 k1 ... k31 s0 s1 05 05 05 05 05The octets k0 to k31 are the session key, and s0 and s1 are the checksum of the session key octets. This encoding lets the sender hide the size of the symmetric encryption key used to encrypt the data 🙈 For example, if an AES algorithm is used for the session key, the sender MAY use 21, 13, and 5 octets of padding for AES-128, AES-192, and AES-256, respectively. That gets everything to the same 40 octets total as the input to the key wrapping method 🦋
In a version 6 Public Key Encrypted Session Key packet, the symmetric algorithm isn't included, as described in Section 5.1. For example, an AES-256 session key would be composed like this:
k0 k1 ... k31 s0 s1 06 06 06 06 06 06Again, k0 to k31 are the session key and s0 and s1 are the checksum. Assuming an AES algorithm is used for the session key, the sender MAY use 22, 14, and 6 octets of padding for AES-128, AES-192, and AES-256, respectively, so the input comes to the same 40 octets total 💜
The output has two fields. The first is the MPI containing the ephemeral key used to establish the shared secret. The second is made of these two subfields:
- One octet encoding the size in octets of the result of the key wrapping method. The value 255 is reserved for future extensions.
- Up to 254 octets representing the result of the key wrapping method, applied to the 8-octet padded session key as described above.
Quick note: for session key sizes of 128, 192, and 256 bits, the size of the key wrapping result is 32, 40, and 48 octets respectively, unless size obfuscation is used 👑
Here's a quick synopsis of the encoding method. The normative sources are still this section, [SP800-56A], and [RFC3394], so the synopsis is just for convenience 🌈
- Obtain the authenticated recipient public key R
- Generate an ephemeral, single-use key pair
{v, V=vG} - Compute the shared point
S = vR m = symm_alg_ID || session key || checksum || pkcs5_paddingcurve_OID_len = (octet)len(curve_OID)Param = curve_OID_len || curve_OID || public_key_alg_ID || 03 || 01 || KDF_hash_ID || KEK_alg_ID for AESKeyWrap || 41 6E 6F 6E 79 6D 6F 75 73 20 53 65 6E 64 65 72 20 20 20 20 || recipient_fingerprintZ_len = the key size for the KEK_alg_ID used with AESKeyWrap- Compute
Z = KDF( S, Z_len, Param ) - Compute
C = AESKeyWrap( Z, m )(per [RFC3394]) - Wipe the memory that contained S, v, and Z to avoid leaking ephemeral secrets 🫣
VB = convert point V to the octet string- Output
(MPI(VB) || len(C) || C)
Decryption is the reverse of all that. The recipient with key pair (r,R) gets the shared secret by calculating:
S = rV = rvG11.5.1. ECDH Parameters
ECDH keys have a hash algorithm parameter for key derivation and a symmetric algorithm for key encapsulation 🎀
For version 6 keys, the following algorithms MUST be used depending on the curve. An implementation MUST NOT generate a version 6 ECDH key over any listed curve that uses different KDF or KEK parameters. An implementation MUST NOT encrypt any message to a version 6 ECDH key over a listed curve that announces a different KDF or KEK parameter. no exceptions, bestie 💀
For version 4 keys, the following algorithms SHOULD be used depending on the curve. An implementation SHOULD only use an AES algorithm as a KEK algorithm 🦋
+==================+================+=====================+
| Curve | Hash Algorithm | Symmetric Algorithm |
+==================+================+=====================+
| NIST P-256 | SHA2-256 | AES-128 |
+------------------+----------------+---------------------+
| NIST P-384 | SHA2-384 | AES-192 |
+------------------+----------------+---------------------+
| NIST P-521 | SHA2-512 | AES-256 |
+------------------+----------------+---------------------+
| brainpoolP256r1 | SHA2-256 | AES-128 |
+------------------+----------------+---------------------+
| brainpoolP384r1 | SHA2-384 | AES-192 |
+------------------+----------------+---------------------+
| brainpoolP512r1 | SHA2-512 | AES-256 |
+------------------+----------------+---------------------+
| Curve25519Legacy | SHA2-256 | AES-128 |
+------------------+----------------+---------------------+
Table 30: OpenPGP ECDH KDF and KEK Parameters Registry12. Notes on Algorithms ✨
12.1. PKCS#1 Encoding in OpenPGP
This specification uses the PKCS#1 functions EME-PKCS1-v1_5 and EMSA-PKCS1-v1_5. The calling conventions of these functions have changed over time, though, so to avoid confusion and interoperability problems, local copies are included in this document, adapted from the ones in PKCS#1 v2.1 [RFC8017]. [RFC8017] is still the ultimate authority on PKCS#1 for OpenPGP, but we think it's worth having a self-contained document that avoids problems later if the conventions change again 🌸
12.1.1. EME-PKCS1-v1_5-ENCODE
Input:
k= key modulus length in octets.M= message to be encoded; an octet string of length mLen, where mLen <= k - 11.
Output:
EM= encoded message; an octet string of length k.
Error: "message too long" 🙈
- Length checking: If mLen > k - 11, output "message too long" and stop.
- Generate an octet string PS of length k - mLen - 3 consisting of pseudorandomly generated non-zero octets. The length of PS will be at least 8 octets.
- Concatenate PS, the message M, and other padding to form an encoded message EM of length k octets as:
EM = 0x00 || 0x02 || PS || 0x00 || M.
- Output EM.
12.1.2. EME-PKCS1-v1_5-DECODE
Input:
EM= encoded message; an octet string.
Output:
M= decoded message; an octet string.
Error: "decryption error" 💅
To decode an EME-PKCS1_v1_5 message, split the encoded message EM into an octet string PS made of non-zero octets and a message M, like this:
EM = 0x00 || 0x02 || PS || 0x00 || M.If the first octet of EM isn't hexadecimal 0x00, the second octet of EM isn't hexadecimal 0x02, there's no octet with hexadecimal value 0x00 to separate PS from M, or the length of PS is less than 8 octets, output "decryption error" and stop. Also see Section 13.5 for how a decryption error is reported differently from a padding error, because that difference matters 🧚
12.1.3. EMSA-PKCS1-v1_5 🦋
okay so this encoding method is totally deterministic 🔮 and it only has an encoding operation. no decoding here, bestie 💅
Input:
Hash: the hash function to be used.M: the message to be encoded.emLen: the intended length of the encoded message in octets, at leasttLen + 11, wheretLenis the octet length of the DER encodingTof a certain value computed during the encoding operation.
Output:
EM: the encoded message, an octet string of lengthemLen.
Errors: "message too long"; "intended encoded message length too short"
Steps:
- Apply the hash function to the message
Mto produce hash valueH:H = Hash(M).- If the hash function outputs "message too long," output "message too long" and stop.
- Let
Tbe the Full Hash Prefix from Table 24 for the given hash function, concatenated with the hash digestH(representing an ASN.1 DER value for the hash function used and the hash digest). LettLenbe the length in octets ofT. - If
emLen<tLen + 11, output "intended encoded message length too short" and stop. - Generate an octet string
PSconsisting ofemLen - tLen - 3octets with hexadecimal value0xFF. The length ofPSwill be at least 8 octets. - Concatenate
PS, the hash prefixT, and other padding to form the encoded messageEMasEM = 0x00 || 0x01 || PS || 0x00 || T. - Output
EM.
12.2. Symmetric Algorithm Preferences 🎀
okay so the symmetric algorithm preference is an ordered list of algorithms that the keyholder accepts 📋 and since it lives on a self-signature, a keyholder can have multiple, different preferences at once 🫶 for example, Alice might have AES-128 only for "alice@work.com" but Camellia-256, Twofish, and AES-128 for "alice@home.org". preferences can also show up in a subkey's binding signature, so keep an eye out 👀
AES-128 is the algorithm that MUST be implemented, so if it isn't explicitly in the list, it's tacitly at the end. it's still good form to put it there explicitly ✨ and if an implementation doesn't implement the preference at all, it's implicitly an AES-128-only implementation 🥺 also, implementations conforming to the previous version of this specification [RFC4880] have TripleDES as the only algorithm that MUST be implemented 🕰️
an implementation MUST NOT use a symmetric algorithm that is not in the recipient's preference list 🚫 when encrypting to more than one recipient, it finds a suitable algorithm by taking the intersection of the preferences of the recipients. and since AES-128 MUST be implemented, that intersection is guaranteed to be non-empty 💖
if an implementation can decrypt a message that a keyholder doesn't have in their preferences, it SHOULD decrypt the message anyway, but it MUST warn the keyholder ⚠️ like, say Alice's implementation supports every algorithm in this spec, but she prefers subsets for work or home. if she gets a message encrypted with IDEA, which isn't in her preferences, the implementation warns her that someone sent an IDEA-encrypted message, but it would ideally decrypt it anyway 🔐
12.2.1. Plaintext 🫧
Algorithm 0, "plaintext", may only be used to denote secret keys that are stored in the clear 🤍 an implementation MUST NOT use algorithm 0 as the indicated symmetric cipher for an encrypted data packet (Sections 5.7 or 5.13). it can use a Literal Data packet (Section 5.9) to encode unencrypted literal data 📄
12.3. Other Algorithm Preferences 🌸
other algorithm preferences work similarly to the symmetric algorithm preference, since they also say which algorithms the keyholder accepts 💅 there are two interesting cases that need extra comments: the compression preferences and the hash preferences 👀
12.3.1. Compression Preferences 🗜️
like the algorithm preferences, an implementation MUST NOT use an algorithm that isn't in the preference vector 🚫 if Uncompressed (0) isn't explicitly in the list, it's tacitly at the end. that means uncompressed messages can always be sent 🍓
heads up though: earlier implementations may assume that no compression preferences means [ZIP(1), Uncompressed(0)] are preferred, and they'll default to ZIP compression 😬 so an implementation that prefers uncompressed data SHOULD explicitly state this in the Preferred Compression Algorithms 🫶
12.3.1.1. Uncompressed 🧊
Algorithm 0, "uncompressed", may only be used to denote a preference for uncompressed data. an implementation MUST NOT use algorithm 0 as the indicated compression algorithm in a Compressed Data packet (Section 5.6). it can use a Literal Data packet (Section 5.9) to encode uncompressed literal data ✨
12.3.2. Hash Algorithm Preferences 🔏
typically the signer picks the hash algorithm, not the verifier, because a signer rarely knows who's going to verify the signature 🤷♀️ this preference lets a protocol based on digital signatures ease in negotiation 🤝
so if Alice is authenticating herself to Bob with a signature, it makes sense for her to use a hash algorithm that Bob's implementation uses. this preference lets Bob state which algorithms Alice may use in his key 👑
since SHA2-256 is the hash algorithm that MUST be implemented, if it isn't explicitly in the list, it's tacitly at the end. again, it's good form to place it there explicitly 💜
12.4. RSA 💎
the PKCS1-v1_5 padding scheme, used by the RSA algorithms defined in this document, is no longer recommended, and [SP800-131A] deprecates its use 😭 so an implementation SHOULD NOT generate RSA keys 🙅♀️
there are algorithm types for RSA Sign-Only and RSA Encrypt-Only keys. these are deprecated in favor of the Key Flags signature subpacket. an implementation MUST NOT create such a key, but it MAY interpret it 🧐
an implementation MUST NOT generate RSA keys of a size less than 3072 bits 🔒 it SHOULD NOT encrypt, sign, or verify using RSA keys of a size less than 3072 bits, and it MUST NOT encrypt, sign, or verify using RSA keys of a size less than 2048 bits 💀 an implementation that decrypts a message using an RSA secret key of a size less than 3072 bits SHOULD generate a deprecation warning that the key is too weak for modern use ⚠️
12.5. DSA 🥀
DSA is no longer recommended, and it's been deprecated in [FIPS186] too 😔 so an implementation MUST NOT generate DSA keys. an implementation also MUST NOT sign or verify using DSA keys 🙅♀️
12.6. Elgamal 🌙
the PKCS1-v1_5 padding scheme, used by the Elgamal algorithm defined in this document, is no longer recommended, and [SP800-131A] deprecates its use 🥲 therefore, an implementation MUST NOT generate Elgamal keys.
an implementation MUST NOT encrypt using Elgamal keys. an implementation that decrypts a message using an Elgamal secret key SHOULD generate a deprecation warning that the key is too weak for modern use 🫠
12.7. EdDSA 🌊
even though the EdDSA algorithm allows arbitrary data as input, using it with OpenPGP requires that a digest of the message be used as input (pre-hashed) 📝 see Section 5.2.4 for details. truncation of the resulting digest is never applied. the resulting digest value is used verbatim as input to the EdDSA algorithm 🔢
for clarity: while [RFC8032] describes different variants of EdDSA, OpenPGP uses the "pure" variant (PureEdDSA) 💫 the hashing that happens with OpenPGP is done as part of the standard OpenPGP signature process, and that hash itself is fed as the input message to the PureEdDSA algorithm 🧁
as specified in [RFC8032], Ed448 also expects a "context string". in OpenPGP, Ed448 is used with the empty string as its context string 🤍
12.8. Reserved Algorithm IDs 🚧
a bunch of algorithm IDs have been reserved for algorithms that would be useful in an OpenPGP implementation 🧩 but there are issues that stop an implementer from actually implementing them. these are marked as reserved in Section 9.1 📌
the reserved public key algorithm X9.42 (21) doesn't have the necessary parameters, parameter order, or semantics defined. the same is currently true for the reserved public key algorithms AEDH (23) and AEDSA (24) 😵💫
previous versions of the OpenPGP specification allowed Elgamal [ELGAMAL] signatures with a public key algorithm ID of 20. those are no longer permitted 🚫 an implementation MUST NOT generate such keys, and it MUST NOT generate Elgamal signatures; see [BLEICHENBACHER] 🔥
12.9. CFB Mode 🌷
the Cipher Feedback (CFB) mode used in this document is defined in Section 6.3 of [SP800-38A] 📖
the CFB segment size s is equal to the block size of the cipher (i.e., n-bit CFB mode, where n is the block size used) 🧮
12.10. Private or Experimental Parameters 🧪
S2K Specifiers, Signature Subpacket Type IDs, User Attribute Subpacket Type IDs, image format IDs, and the various algorithm IDs described in Section 9 all reserve the range 100 to 110 for Private and Experimental Use 🎈 Packet Type IDs reserve the range 60 to 63 for Private and Experimental Use too. these are intentionally managed by the Private Use and Experimental Use policies, as described in [RFC8126] 🌈
however, implementations need to be careful with these. they should promote them to full IANA-managed parameters when they grow beyond the original, limited system 🌱
12.11. Meta Considerations for Expansion 🔭
if OpenPGP gets extended in a way that isn't backward compatible (meaning old implementations won't gracefully handle the absence of a new feature), the extension proposal can be declared in the keyholder's self-signature as part of the Features signature subpacket 🧑💻
we can't say definitively which extensions won't be forward compatible 🤔 but typically new algorithms are forward compatible, whereas new packets are not 📦
if an extension proposal doesn't update the Features system, it SHOULD include an explanation of why that's unnecessary. if the proposal contains neither an extension to the Features system nor an explanation of why such an extension is unnecessary, the proposal SHOULD be rejected 🙅♀️💅
13. Security Considerations 🔐✨
- Okay bestie, cryptography is a whole minefield, so implementers should check the current literature to see if any algorithms used here have been found to be vulnerable to an attack. If they have, implementers should consider disallowing those algorithms for new data and warning the end user, or preventing use, when someone is trying to consume data protected by algorithms that are now vulnerable. 🫶
- This spec runs on Public Key Cryptography 🔑. It's assumed that the private key half of a public-private key pair is controlled and secured by the proper party or parties. Guard that private key like your life depends on it, because it kind of does. 💅
- MD5 and SHA-1 are giving "broken" energy 💀. Collisions have been found in a number of cases. MD5 and SHA-1 are deprecated for use in OpenPGP (see Section 9.5). Let them go, periodt.
- Lots of security protocol designers think it's a bad idea to use a single key for both privacy (encryption) and integrity (signatures) 🙈. Honestly this was one of the motivating forces behind the version 4 key format with separate signature and encryption keys. Using a single key for encrypting and signing is discouraged.
- The DSA algorithm will work with any hash, but it is sensitive to the quality of the hash algorithm 🧪. Verifiers should be aware that even if the signer used a strong hash, an attacker could have modified the signature to use a weak one. Only signatures using acceptably strong hash algorithms should be accepted as valid.
- OpenPGP mixes a bunch of asymmetric, symmetric, and hash algorithms, and each one has its own strength level 💪. So care should be taken to ensure that the weakest element of an OpenPGP Message is still sufficiently strong for the purpose at hand. Consensus about the strength of a given algorithm may evolve, but NIST Special Publication 800-57 [SP800-57] contains recommendations (current at the time of this publication) about equivalent security levels of different algorithms 📚.
- There's a somewhat-related potential security problem in signatures, and it's a bit of a plot twist 😱. If an attacker can find a message that hashes to the same hash with a different algorithm, a bogus signature structure can be constructed that evaluates correctly.
For example, suppose Alice DSA-signs message M using hash algorithm H. Suppose that Mallet finds a message M' that has the same hash value as M with H'. Mallet can then construct a signature block that verifies as Alice's signature of M' with H'. However, this would also constitute a weakness in either H or H', or both. Should this ever occur, a revision will have to be made to this document to revise the allowed hash algorithms 🚨.
- If you're building an authentication system, the recipient may specify a preferred signing algorithm. However, the signer would be foolish to use a weak algorithm simply because the recipient requests it. Don't be that girl 🙅♀️.
- Some of the encryption algorithms mentioned in this document have been analyzed less than others. For example, although TWOFISH is presently considered reasonably strong, it has been analyzed much less than AES. Other algorithms may have other concerns surrounding them 🦋.
- In late summer 2002, Jallad, Katz, and Schneier published an interesting attack on previous versions of the OpenPGP specification and some of its implementations [JKS02]. The attacker modifies a message and sends it to a user, who then returns the erroneously decrypted message to the attacker. The attacker is thus using the user as a decryption oracle and can often decrypt the message 😭. This attack is a particular form of ciphertext malleability. See Section 13.7 for information on how to defend against such an attack using more recent versions of OpenPGP 🛡️.
13.1. SHA-1 Collision Detection
As described in [SHAMBLES], the SHA-1 digest algorithm is not collision resistant 💔. However, an OpenPGP implementation cannot completely discard the SHA-1 algorithm, because it is required for implementing version 4 public keys. In particular, the version 4 fingerprint derivation uses SHA-1. So as long as an OpenPGP implementation supports version 4 public keys, it will need to implement SHA-1 in at least some scenarios 🥲.
To avoid the risk of uncertain breakage from a maliciously introduced SHA-1 collision, an OpenPGP implementation MAY attempt to detect when a hash input is likely from a known collision attack, and then either reject the hash input deliberately or modify the hash output 🧠. This should convert an uncertain breakage (where it is unclear what the effect of a collision will be) to an explicit breakage, which is more desirable for a robust implementation ✨.
[STEVENS2013] describes a method for detecting indicators of well-known SHA-1 collision attacks. Some example C code implementing this technique can be found at [SHA1CD] 💻.
13.2. Advantages of Salted Signatures
Version 6 signatures include a salt that is hashed first, and its size depends on the hashing algorithm 🧂. This makes version 6 OpenPGP signatures non-deterministic and protects against a broad class of attacks that depend on creating a signature over a predictable message. By selecting a new random salt for each signature made, the signed hashes and the signatures are not predictable 🎲.
While the material to be signed could be attacker controlled, hashing the salt first means that there is no attacker-controlled hashed prefix. An example of this kind of attack is described in the paper "SHA-1 is a Shambles" [SHAMBLES], which leverages a chosen prefix collision attack against SHA-1. This means that an adversary carrying out a chosen-message attack will not be able to control the hash that is being signed, and will need to break second-preimage resistance instead of the simpler collision resistance to create two messages having the same signature. The size of the salt is bound to the hash function to match the expected collision-resistance level, and it is at least 16 octets 🔒.
In some cases, an attacker may be able to induce a signature to be made, even if they do not control the content of the message. In some scenarios, a repeated signature over the exact same message may risk leakage of part or all of the signing key. For example, see the discussion of hardware faults over EdDSA and deterministic ECDSA in [PSSLR17] 😬. Choosing a new random salt for each signature ensures that no repeated signatures are produced, which mitigates this risk 🌸.
13.3. Elliptic Curve Side Channels
Side-channel attacks are a concern when a compliant application's use of the OpenPGP format can be modeled by a decryption or signing oracle. For example, this happens when an application is a network service performing decryption for unauthenticated remote users 👀. ECC scalar multiplication operations used in ECDSA and ECDH are vulnerable to side-channel attacks. Countermeasures can often be taken at the higher protocol level, such as limiting the number of allowed failures or time-blinding the operations associated with each network interface ⏱️. Mitigations at the scalar multiplication level seek to eliminate any measurable distinction between the ECC point addition and doubling operations 🧚.
13.4. Risks of a Quick Check Oracle
In winter 2005, Serge Mister and Robert Zuccherato from Entrust released a paper describing a way that the "quick check" in v1 SEIPD and SED packets can be used as an oracle to decrypt two octets of every cipher block [MZ05]. This check was intended for early detection of session key decryption errors, particularly to detect a wrong passphrase, since v4 SKESK packets do not include an integrity check 🫣.
There is a danger when using the quick check if timing or error information about the check can be exposed to an attacker, particularly via an automated service that allows rapidly repeated queries. 🚩
Disabling the quick check prevents the attack 🛑.
For very large encrypted data whose session key is protected by a passphrase using a v4 SKESK, the quick check may be convenient to the user, since it informs them early that they typed the wrong passphrase. But the implementation should use the quick check with care. The recommended approach for secure and early detection of decryption failure is to encrypt data using v2 SEIPD 🎀. If the session key is public key encrypted, the quick check is not useful, because the public key encryption of the session key should guarantee that it is the right session key 🔑.
The quick check oracle attack is a particular type of attack that exploits ciphertext malleability. For information about other similar attacks, see Section 13.7 🔁.
13.5. Avoiding Leaks from PKCS#1 Errors
The PKCS#1 padding (used in RSA-encrypted and ElGamal-encrypted PKESK) has been found to be vulnerable to attacks. In these attacks, a system that allows distinguishing padding errors from other decryption errors can act as a decryption and/or signing oracle, which can leak the session key or allow signing arbitrary data, respectively [BLEICHENBACHER-PKCS1]. The number of queries required to carry out an attack can range from thousands to millions, depending on how strict and careful an implementation is in processing the padding 😱.
To make the attack more difficult, an implementation SHOULD implement strict, robust, and constant time padding checks 🧱.
To prevent the attack, in settings where the attacker does not have access to timing information concerning message decryption, the simplest solution is to report a single error code for all variants of PKESK processing errors, as well as SEIPD integrity errors. This also includes session key parsing errors, such as an invalid cipher algorithm for v3 PKESK, or a session key size mismatch for v6 PKESK 🧾. If the attacker may have access to timing information, then a constant time solution is also needed. This requires careful design, especially for v3 PKESK, where session key size and cipher information is typically not known in advance, because it is part of the PKESK encrypted payload 🧩.
13.6. Fingerprint Usability 💅
Okay bestie, so this spec uses fingerprints in a bunch of places on the wire (like Sections 5.2.3.23, 5.2.3.35, and 5.2.3.36) and in processing too (like the ECDH KDF in Section 11.5) 🧠. An implementation can also use the fingerprint internally, like as an index into a keystore 🔑.
Now, some OpenPGP users have historically compared fingerprints by hand to verify a peer's public key 👀. For a version 4 fingerprint, that usually means 40 hexadecimal digits, often chunked into groups of four with whitespace in between 🌸.
ngl, when a human is actively involved, the result is kinda dubious 💀. There's not much evidence that most humans are good at precisely comparing high-entropy data, especially when it's squished into compact text like hex [USENIX-STUDY] 🙈.
The version 6 fingerprint makes things even harder for the human verifier 😭. At 256 bits (compared to version 4's 160-bit fingerprint), a version 6 fingerprint is even harder for a human to successfully compare.
An OpenPGP implementation should prioritize mechanical fingerprint transfer and comparison where possible and SHOULD NOT promote manual transfer or comparison of full fingerprints by a human unless there is no other way to achieve the desired result. Machines are simply better at this 🤖✨.
While this subsection acknowledges existing practice for human-representable version 4 fingerprints, this document does not attempt to standardize any specific human-readable form of version 6 fingerprint for this discouraged use case 🚫.
NOTE: the topic of interoperable human-in-the-loop key verification needs more work, which will be done in a separate document. 📝
13.7. Avoiding Ciphertext Malleability 🎀
Okay so, if ciphertext can be modified by an attacker but still gets decrypted to some new plaintext, it's considered "malleable" 😈. A number of attacks can show up in any cryptosystem that uses malleable encryption, so [RFC4880] and later versions of OpenPGP offer mechanisms to defend against it 🛡️. The catch is that OpenPGP data may have been created before these defenses existed. Since OpenPGP implementations deal with historic stored data, they may run into malleable ciphertexts 📦.
When an OpenPGP implementation discovers that it is decrypting data that appears to be malleable, it MUST generate a clear error message that indicates the integrity of the message is suspect 🚨. It SHOULD NOT attempt to parse nor release decrypted data to the user, and it SHOULD halt with an error 🛑. Parsing or releasing decrypted data before its integrity has been confirmed can leak the decrypted data [EFAIL] [MRLG15] 😱.
In the case of AEAD encrypted data, if the authentication tag fails to verify, the implementation MUST NOT attempt to parse nor release decrypted data to the user, and it MUST halt with an error 🚫💅.
An implementation that encounters malleable ciphertext MAY choose to release cleartext to the user if it is not encrypted using AEAD, it is known to be dealing with historic archived legacy data, and the user is aware of the risks 🗂️. Use your judgment, babe 🫶.
In the case of AEAD encrypted messages, if the message is truncated, i.e., the final zero-octet chunk and possibly (part of) some chunks before it are missing, the implementation MAY choose to release cleartext from the fully authenticated chunks before it to the user if it is operating in a streaming fashion. However, it MUST indicate a clear error message as soon as the truncation is detected ✂️.
Any of the following OpenPGP data elements indicate that malleable ciphertext is present 👀:
- All Symmetrically Encrypted Data packets (Section 5.7).
- Within any encrypted container, any Compressed Data packet (Section 5.6) where there is a decompression failure.
- Any version 1 Symmetrically Encrypted and Integrity Protected Data packet (Section 5.13.1) where the internal Modification Detection Code does not validate.
- Any version 2 Symmetrically Encrypted and Integrity Protected Data packet (Section 5.13.2) where the authentication tag of any chunk fails or where there is no final zero-octet chunk.
- Any Secret-Key packet with encrypted secret key material (Section 3.7.2.1) where there is an integrity failure, based on the value of the secret key protection octet:
- Value 253 (AEAD): where the AEAD authentication tag is invalid.
- Value 254 (CFB): where the SHA1 checksum is mismatched.
- Value 255 (MalleableCFB) or raw cipher algorithm: where the trailing 2-octet checksum does not match.
To avoid these circumstances, an implementation that generates OpenPGP encrypted data SHOULD select the encrypted container format with the most robust protections that can be handled by the intended recipients 🏆. In particular:
- The SED packet is deprecated and MUST NOT be generated. Bye SED 👋💀.
- When encrypting to one or more public keys:
- If all recipient keys indicate support for a version 2 Symmetrically Encrypted and Integrity Protected Data packet in their Features signature subpacket (Section 5.2.3.32), if all recipient keys are version 6 keys without a Features signature subpacket, or the implementation can otherwise infer that all recipients support v2 SEIPD packets, the implementation SHOULD encrypt using a v2 SEIPD packet ✨.
- If one of the recipients does not support v2 SEIPD packets, then the message generator MAY use a v1 SEIPD packet instead 🔄.
- Passphrase-protected secret key material in a version 6 Secret Key or version 6 Secret Subkey packet SHOULD be protected with AEAD encryption (S2K usage octet 253) unless it will be transferred to an implementation that is known to not support AEAD 🔐. An implementation should be aware that, in scenarios where an attacker has write access to encrypted private keys, CFB-encrypted keys (S2K usage octet 254 or 255) are vulnerable to corruption attacks that can cause leakage of secret data when the secret key is used [KOPENPGP] [KR02] 😭🔓.
Implementers should implement AEAD (v2 SEIPD and S2K usage octet 253) promptly and encourage its spread 📣💜.
Users are RECOMMENDED to migrate to AEAD 🌈🫶.
13.8. Secure Use of the v2 SEIPD Session-Key-Reuse Feature 🦋
Okay, the salted key derivation of v2 SEIPD packets (Section 5.13.2) lets the recipient of an encrypted message reply to the sender and all the other recipients without needing their public keys 🤯. They just use the same v6 PKESK packets they received, plus a different random salt value. This is a secure mechanism on the cryptographic level that makes message encryption possible when a sender doesn't have a copy of an encryption-capable certificate for one or more participants in the conversation, so it can seriously level up the overall security of an application 📈💖. However, care must be taken when using this mechanism not to create security vulnerabilities, such as the following 🚨:
- Replying to only a subset of the original recipients and the original sender by use of the session-key-reuse feature would mean that the remaining recipients (including the sender) of the original message could read the encrypted reply message, too 👀.
- Adding a further recipient to the reply that is encrypted using the session-key-reuse feature gives that further recipient cryptographic access to the original message being replied to (and potentially to a longer history of previous messages) 😬.
- A modification of the list of recipients addressed in the above points also needs to be safeguarded when a message is initially composed as a reply with session-key reuse, but then is stored (e.g., as a draft) and later reopened for further editing and to be finally sent 📝.
- There is the potential threat that an attacker with network or mailbox access, who is at the same time a recipient of the original message, silently removes themselves from the message before the victim's client receives it 😈. The victim's client that then uses the mechanism for replying with session-key reuse would unknowingly compose an encrypted message that could be read by the attacker 💀. Implementations are encouraged to use the Intended Recipient Fingerprint subpacket (Section 5.2.3.36) when composing messages and checking the consistency of the set of recipients of a message before replying to it with session-key reuse 🔍.
- When using the session-key-reuse feature in any higher-layer protocol, care should be taken to ensure that there is no other potentially interfering practice of session-key reuse established in that protocol. Such interfering session-key reuse could, for instance, be given (if an initial message is already composed) by reusing the session key of an existing encrypted file that may have been shared among a group of users already. Using the session-key-reuse feature to compose an encrypted reply to such a message would unknowingly give this whole group of users cryptographic access to the encrypted message 🫣.
- Generally, the use of the session-key-reuse feature should be under the control of the user. Specifically, care should be taken so that this feature is not silently used when the user assumes that proper public key encryption is used. This can be the case, for instance, when the public key of one of the recipients of the reply is known but has expired ⏰. Special care should be taken to ensure that users do not get caught in continued use of the session-key reuse unknowingly, but instead receive the chance to switch to proper fresh public key encryption as soon as possible 🙏.
- Whenever possible, a client should prefer a fresh public key encryption over the session-key reuse 🥇.
Even though this is not necessarily a security aspect, note that initially composing an encrypted reply using the session-key-reuse feature on one client, then storing it (e.g., as a draft) and later reopening the stored unfinished reply with another client that does not support the session-key-reuse feature, may lead to interoperability problems 🧩.
Avoiding the pitfalls described above requires context-specific expertise 🧠💅. An implementation should only make use of the session-key-reuse feature in any particular application layer when it can follow reasonable documentation about how to deploy the feature safely in that specific application. At the time of this writing, there is no known documentation about safe reuse of OpenPGP session keys for any specific context 📚. An implementer that intends to make use of this feature should publish their own proposed guidance for others to review 🌸✨.
13.9. Escrowed Revocation Signatures 🔐
Okay so, Alice is the keyholder 💖 and she might want to hand a third party the power to revoke her key.
The fave way for Alice to do this is to make a specific Revocation Signature (Signature Type ID 0x20, 0x28, or 0x30) and send it securely to a preferred revoker, who holds it in escrow. Then the preferred revoker can publish that escrowed signature whenever it makes sense, instead of making a brand new Revocation Signature themselves 🦋
There are a bunch of perks to escrowed signatures over the deprecated Revocation Key subpacket (Section 5.2.3.23) ✨
- The keyholder can limit what kinds of revocation the preferred revoker can issue, just by escrowing only those specific signatures 🎀
- There's no public or visible link between the keyholder and the preferred revoker 🤫
- Third parties can verify the revocation without needing to find the preferred revoker's key 💅
- The preferred revoker doesn't even need a public OpenPGP Key if some other secure transport is possible between them and the keyholder 🫶
- Implementation support for enforcing a revocation from an authorized Revocation Key subpacket is uneven and unreliable 💀
- If the fingerprint mechanism suffers a cryptanalytic flaw, the escrowed Revocation Signature is not affected 👑
A Revocation Signature can also be split into shares and spread across multiple parties, so some subset of those parties has to team up before the escrowed Revocation Signature gets recreated 🧚
13.10. Random Number Generation and Seeding 🎲
OpenPGP requires a cryptographically secure pseudorandom number generator (CSPRNG). In most cases the operating system has an appropriate facility, like the getrandom() syscall on Linux or BSD, which should be used unless other concerns (like performance) come up. It is RECOMMENDED to use an existing CSPRNG implementation instead of crafting a new one. Lots of adequate cryptographic libraries are already out there under favorable license terms. If those prove unsatisfactory, [RFC4086] has guidance on generating random values 🌸
OpenPGP uses random data with three different levels of visibility 👀
- In publicly visible fields such as nonces, IVs, public padding material, or salts.
- In shared-secret values, such as session keys for encrypted data or padding material within an encrypted packet.
- In entirely private data, such as asymmetric key generation.
With a properly functioning CSPRNG, this range of visibility is not a security problem, because it's not feasible to determine the CSPRNG state from its output. However, with a broken CSPRNG, an attacker might use visible output to figure out the CSPRNG internal state and then predict less-visible data like keying material, as documented in [CHECKOWAY]. 😱
An implementation can add extra security against this kind of attack by using separate CSPRNGs to generate random data with different levels of visibility 🧠
13.11. Traffic Analysis 🕵️♀️
When OpenPGP data travels through the network, the size of the data can leak information to an attacker. There are circumstances where that leak could be unacceptable from a security perspective 🙈
For example, say possible cleartext messages for a given protocol are known to be either yes (3 octets) or no (2 octets), and they're sent within a Symmetrically Encrypted and Integrity Protected Data packet. The length of the encrypted message will reveal the contents of the cleartext. Yikes 💀
Another one: sending an OpenPGP Transferable Public Key over an encrypted network connection might reveal the length of the certificate. Since the length of an OpenPGP certificate varies based on content, an external observer who's interested in metadata (e.g., which individual is trying to contact another individual) might be able to guess the identity of the certificate sent, if its length is unique 🔍
In both cases, an implementation can adjust the size of the compound structure by including a Padding packet (see Section 5.14) 🎀
13.12. Surreptitious Forwarding 📨
When an attacker obtains a signature for some text, e.g., by receiving a signed message, they may be able to use that signature maliciously. They could send a message pretending to come from the original sender, with the same body and signature, to a different recipient. To prevent this, an implementation SHOULD implement the Intended Recipient Fingerprint subpacket (Section 5.2.3.36) 🛡️
13.13. Hashed vs. Unhashed Subpackets 🧁
Each OpenPGP signature can have subpackets in two different sections. The first set (the "hashed section") is covered by the signature itself. The second set has no cryptographic protections and is used for advisory material only, including locally stored annotations about the signature 📝
For example, consider an implementation working with a specific signature that happens to know the signature was made by a certain key, even though the signature has no Issuer Fingerprint subpacket (Section 5.2.3.35) in the hashed section. That implementation MAY synthesize an Issuer Fingerprint subpacket and store it in the unhashed section, so it can recall which key issued the signature in the future ✨
Some subpackets are only useful when they're in the hashed section, and an implementation SHOULD ignore them when they're found with unknown provenance in the unhashed section. For example, a Preferred AEAD Ciphersuites subpacket (Section 5.2.3.15) in a Direct Key self-signature indicates the keyholder's preferences when encrypting v2 SEIPD data to the key. An implementation that observes such a subpacket in the unhashed section would open itself to an attack where the recipient's certificate is tampered with to push a specific cipher or mode of operation. Not okay 🚨
13.14. Malicious Compressed Data 🗜️
It is possible to form a compression quine that produces itself upon decompression. That leads to infinite regress in any implementation willing to parse arbitrary numbers of layers of compression. This could cause resource exhaustion, which itself could lead to termination by the operating system. If the operating system creates a "crash report", that report could contain confidential information 😭
An OpenPGP implementation SHOULD limit the number of layers of compression it is willing to decompress in a single message 🫶
14. Implementation Considerations 🛠️
This section is a collection of comments to help an implementer who has a particular interest in backward compatibility. Often the differences are small, but small differences are frequently more vexing than large ones. So this is a non-comprehensive list of potential problems and gotchas for a developer trying to achieve backward compatibility 🧚
- There are many possible ways for two keys to have the same key material but different fingerprints (and thus different Key IDs). For example, since a version 4 fingerprint is constructed by hashing the key creation time along with other things, two version 4 keys created at different times yet with the same key material will have different fingerprints 🔍
- OpenPGP does not put limits on the size of public keys. However, larger keys are not necessarily better keys. Larger keys take more computation time to use, and this can quickly become impractical. Different OpenPGP implementations may also use different upper bounds for public key sizes, so care should be taken when choosing sizes to maintain interoperability 🐢
- ASCII Armor is an optional feature of OpenPGP. The OpenPGP Working Group strives for a minimal set of mandatory-to-implement features, and since there could be useful implementations that only use binary object formats, this is not a "MUST" feature for an implementation. For example, an implementation that is using OpenPGP as a mechanism for file signatures may find ASCII Armor unnecessary. OpenPGP permits an implementation to declare what features it does and does not support, but ASCII Armor is not one of these. Since most implementations allow binary and armored objects to be used indiscriminately, an implementation that does not implement ASCII Armor may find itself with compatibility issues with general-purpose implementations. Moreover, implementations of OpenPGP-MIME [RFC3156] already have a requirement for ASCII Armor, so those implementations will necessarily have support 🎀
- What this document calls the "Legacy packet format" (Section 4.2.2) is what older documents called the "old packet format". It is the packet format used by implementations predating [RFC2440]. The current "OpenPGP packet format" (Section 4.2.1) was called the "new packet format" by older RFCs. This is the format introduced in [RFC2440] and maintained through [RFC4880] to this document 📜
14.1. Constrained Legacy Fingerprint Storage for Version 6 Keys 📦
Some OpenPGP implementations have fixed length constraints for key fingerprint storage that will not fit all 32 octets of a version 6 fingerprint. For example, [OPENPGPCARD] reserves 20 octets for each stored fingerprint 🧩
An OpenPGP implementation MUST NOT attempt to map any part of a version 6 fingerprint to such a constrained field unless the relevant specification for the constrained environment has explicit guidance for storing a version 6 fingerprint that distinguishes it from a version 4 fingerprint. An implementation interacting with such a constrained field SHOULD directly calculate the version 6 fingerprint from public key material and associated metadata instead of relying on the constrained field 🚫
15. IANA Considerations 📋
This document obsoletes [RFC4880]. IANA has updated all registration information that references [RFC4880] to reference this RFC instead 💜
15.1. Renamed Protocol Group 🌈
IANA bundles a set of registries associated with a particular protocol into a "protocol group". IANA has updated the name of the "Pretty Good Privacy (PGP)" protocol group (i.e., the group of registries described at <https://www.iana.org/assignments/pgp-parameters>) to "OpenPGP". IANA has arranged a permanent redirect from the existing URL to the new URL for the registries in this protocol group. All further updates specified below are for registries within this same OpenPGP protocol group 🦋
15.2. Renamed and Updated Registries ✨
okay so IANA gave a bunch of registries a glow-up 💅 They renamed the "PGP String-to-Key (S2K)" registry to OpenPGP String-to-Key (S2K) Types and refreshed what's inside it, as shown in Table 1. 🎀
They also renamed the "PGP Packet Types/Tags" registry to OpenPGP Packet Types and updated it as shown in Table 3. 🦋
The "Signature Subpacket Types" registry is now OpenPGP Signature Subpacket Types, updated as shown in Table 5. 🌸
"Key Server Preference Extensions" became OpenPGP Key Server Preference Flags, with contents updated as shown in Table 8. ✨
"Key Flags Extensions" is now OpenPGP Key Flags, updated as shown in Table 9. 💖
"Reason for Revocation Extensions" is now OpenPGP Reason for Revocation (Revocation Octet), updated as shown in Table 10. 🙈
"Implementation Features" got renamed to OpenPGP Features Flags, with contents updated as shown in Table 11. 🧚
"PGP User Attribute Types" is now OpenPGP User Attribute Subpacket Types, updated as shown in Table 13. 👑
"Image Format Subpacket Types" became OpenPGP Image Attribute Encoding Format, updated as shown in Table 15. 🌈
"Public Key Algorithms" is now OpenPGP Public Key Algorithms, updated as shown in Table 18. 🔥
"Symmetric Key Algorithms" became OpenPGP Symmetric Key Algorithms, updated as shown in Table 21. 💜
"Compression Algorithms" is now OpenPGP Compression Algorithms, updated as shown in Table 22. 🍓
And finally "Hash Algorithms" became OpenPGP Hash Algorithms, updated as shown in Table 23. ⭐
15.3. Removed Registry
IANA marked the empty "New Packet Versions" registry as OBSOLETE. 💀 It's gone, bestie. 🫶
They also added a tombstone note to the OpenPGP protocol group, and it says:
| Those wishing to use the removed "New Packet Versions" registry
| should instead register new versions of the relevant packets in
| the "OpenPGP Key and Signature Versions", "OpenPGP Key IDs and
| Fingerprints", and "OpenPGP Encrypted Message Packet Versions"
| registries.So if you want a new packet version, go register it in one of those three registries instead. 🦋
15.4. Added Registries
IANA also added some brand-new registries to the OpenPGP protocol group. ✨ The initial contents of each one live in the table listed next to it.
OpenPGP Secret Key Encryption (S2K Usage Octet)(Table 2). 🎀OpenPGP Signature Types(Table 4). 💅OpenPGP Signature Notation Data Subpacket Notation Flags(Table 6). 🌸OpenPGP Signature Notation Data Subpacket Types(Table 7). 🧚OpenPGP Key IDs and Fingerprints(Table 12). 🔥OpenPGP Image Attribute Versions(Table 14). 💖OpenPGP Armor Header Lines(Table 16). 🌈OpenPGP Armor Header Keys(Table 17). 🍓OpenPGP ECC Curve OIDs and Usage(Table 19). ⭐OpenPGP ECC Curve-Specific Wire Formats(Table 20). 👑OpenPGP Hash Algorithm Identifiers for RSA Signatures' Use of EMSA-PKCS1-v1_5 Padding(Table 24). 💜OpenPGP AEAD Algorithms(Table 25). 🦋OpenPGP Encrypted Message Packet Versions(Table 26). ✨OpenPGP Key and Signature Versions(Table 27). 🙈OpenPGP Elliptic Curve Point Wire Formats(Table 28). 🌸OpenPGP Elliptic Curve Scalar Encodings(Table 29). 💅OpenPGP ECDH KDF and KEK Parameters(Table 30). 🫶
15.5. Registration Policies
Every registry in the OpenPGP protocol group uses the Specification Required registration policy, except the ones listed in Section 15.5.1. 🎀 See Section 4.6 of [RFC8126] for the details. 💅
This policy means review and approval by a designated expert is required. It also means the IDs and their meanings must be written down in a permanent, readily available public specification, with enough detail that independent implementations can interoperate. 🫶
15.5.1. Registries That Use RFC Required
These registries use the RFC Required registration policy, as described in Section 4.7 of [RFC8126]: 🌈
OpenPGP Packet Types(Table 3). 🔥OpenPGP Key IDs and Fingerprints(Table 12). 💖OpenPGP Encrypted Message Packet Versions(Table 26). 🦋OpenPGP Key and Signature Versions(Table 27). ✨
15.6. Designated Experts
Designated experts check whether new registrations keep the security properties that the base implementation expects. 🛡️ They also check that the new registrations don't cause interoperability issues with existing implementations. The one exception is that it's fine if old implementations don't produce or consume the IDs tied to the new registrations. 💜 If a proposal doesn't pass these checks, it could instead become a new work item for the OpenPGP Working Group or its successor. 🧚
The subsections below give specific guidance for classes of registry updates that a designated expert will look at. 👑
Designated experts should also consider Section 12.11 when they review proposed additions to the OpenPGP protocol group. 🌸
15.6.1. Key and Signature Versions
When someone defines a new version of OpenPGP Keys or Signatures, the OpenPGP Key and Signature Versions registry (Table 27) should be updated. 💅 And when a new version of OpenPGP Key is defined, the OpenPGP Key IDs and Fingerprints registry (Table 12) should also be updated. 🫶
15.6.2. Encryption Versions
When defining a new version of the Symmetrically Encrypted and Integrity Protected Data packet (Section 5.13), Public Key Encrypted Session Key packet (Section 5.1), and/or Symmetric Key Encrypted Session Key packet (Section 5.3), the OpenPGP Encrypted Message Packet Versions registry (Table 26) should be updated. ✨ When the SEIPD is updated, consider also adding a corresponding flag to the OpenPGP Features Flags registry (Table 11). 🌈
15.6.3. Algorithms
Section 9 lists the cryptographic and compression algorithms OpenPGP uses. 🔥 Adding new algorithms is usually simple. Sometimes all you need is to allocate an ID and point to a reference. 💖 But some algorithm registries need some subtle extra details when a new algorithm shows up. 🙈
15.6.3.1. Elliptic Curve Algorithms
To register a new elliptic curve for OpenPGP, you need to do three things. 🦋 Its OID goes in the OpenPGP ECC Curve OIDs and Usage registry (Table 19). Its wire format gets documented in the OpenPGP ECC Curve-Specific Wire Formats registry (Table 20). And if it's used for ECDH, its KDF and KEK parameters must be filled in the OpenPGP ECDH KDF and KEK Parameters registry (Table 30). 💜
If the wire format(s) you use isn't already defined in the OpenPGP Elliptic Curve Point Wire Formats (Table 28) or OpenPGP Elliptic Curve Scalar Encodings (Table 29) registries, it should be defined in those as well. 🎀
15.6.3.2. Symmetric Key Algorithms
When registering a new symmetric cipher with a block size of 64 or 128 bits and a key size that's a multiple of 64 bits, no new considerations are needed. 🌸
If the new cipher has a different block size, there needs to be extra documentation describing how to use the cipher in CFB mode. 🧚
If the new cipher has an unusual key size, then padding needs to be considered for X25519 and X448 key wrapping. Right now, those need no padding. 👑
15.6.3.3. Hash Algorithms
When registering a new hash algorithm in the OpenPGP Hash Algorithms registry (Table 23), if the algorithm is also going to be used with RSA signing schemes, it must also have an entry in the OpenPGP Hash Algorithm Identifiers for RSA Signatures' Use of EMSA-PKCS1-v1_5 Padding registry (Table 24). 💅🔥
16. References ✨
16.1. Normative References 📚✨
okay bestie, this is the normative references list 🫶 these are the docs the spec leans on to work, so if you want the full picture, these are your homework 💅 i've kept every title, number, DOI, and link exactly as written, cause one wrong character and the whole thing breaks 😭
- [AES] NIST, "Advanced Encryption Standard (AES)", Updated May 2023, FIPS PUB 197, DOI 10.6028/NIST.FIPS.197-upd1, November 2001,
<https://nvlpubs.nist.gov/nistpubs/FIPS/NIST.FIPS.197-upd1.pdf>. 🦋 - [BLOWFISH] Schneier, B., "Description of a New Variable-Length Key, 64-Bit Block Cipher (Blowfish)", Fast Software Encryption, Cambridge Security Workshop Proceedings, pp. 191-204, December 1993,
<https://www.schneier.com/academic/archives/1994/09/description_of_a_new.html>. 🌸 - [BZ2] bzip2, "bzip2 and libbzip2", 2010,
<https://sourceware.org/bzip2/>. 🍓 - [EAX] Bellare, M., Rogaway, P., and D. Wagner, "A Conventional Authenticated-Encryption Mode", April 2003,
<https://seclab.cs.ucdavis.edu/papers/eax.pdf>. 💜 - [ELGAMAL] Elgamal, T., "A Public Key Cryptosystem and a Signature Scheme Based on Discrete Logarithms", IEEE Transactions on Information Theory, Vol. 31, Issue 4, pp. 469-472, DOI 10.1109/TIT.1985.1057074, July 1985,
<https://doi.org/10.1109/TIT.1985.1057074>. 🧚 - [FIPS180] NIST, "Secure Hash Standard (SHS)", FIPS PUB 180-4, DOI 10.6028/NIST.FIPS.180-4, August 2015,
<https://nvlpubs.nist.gov/nistpubs/fips/nist.fips.180-4.pdf>. ⭐ - [FIPS186] NIST, "Digital Signature Standard (DSS)", FIPS PUB 186-5, DOI 10.6028/NIST.FIPS.186-5, February 2023,
<https://nvlpubs.nist.gov/nistpubs/FIPS/NIST.FIPS.186-5.pdf>. 🔥 - [FIPS202] NIST, "SHA-3 Standard: Permutation-Based Hash and Extendable-Output Functions", FIPS PUB 202, DOI 10.6028/NIST.FIPS.202, August 2015,
<https://nvlpubs.nist.gov/nistpubs/fips/nist.fips.202.pdf>. 🌈 - [IDEA] Lai, X. and J. L. Massey, "A Proposal for a New Block Encryption Standard", Advances in Cryptology - EUROCRYPT '90, Vol. 473, pp. 389-404, DOI 10.1007/3-540-46877-3_35, January 1991,
<https://link.springer.com/chapter/10.1007/3-540-46877-3_35>. 🎀 - [ISO10646] ISO, "Information technology - Universal coded character set (UCS)", ISO/IEC 10646:2020, December 2020,
<https://www.iso.org/standard/76835.html>. 🌸 - [JFIF] ITU-T, "Information technology - Digital compression and coding of continuous-tone still images: JPEG File Interchange Format (JFIF)", Recommendation ITU-T T.871, May 2011,
<https://www.itu.int/rec/T-REC-T.871-201105-I>. 🦋 - [RFC1321] Rivest, R., "The MD5 Message-Digest Algorithm", RFC 1321, DOI 10.17487/RFC1321, April 1992,
<https://www.rfc-editor.org/info/rfc1321>. 🍓 - [RFC1950] Deutsch, P. and J. Gailly, "ZLIB Compressed Data Format Specification version 3.3", RFC 1950, DOI 10.17487/RFC1950, May 1996,
<https://www.rfc-editor.org/info/rfc1950>. 💜 - [RFC1951] Deutsch, P., "DEFLATE Compressed Data Format Specification version 1.3", RFC 1951, DOI 10.17487/RFC1951, May 1996,
<https://www.rfc-editor.org/info/rfc1951>. 🧚 - [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate Requirement Levels", BCP 14, RFC 2119, DOI 10.17487/RFC2119, March 1997,
<https://www.rfc-editor.org/info/rfc2119>. 👑 this one is the big one, the source of all those capital-letter key words we keep tracking 😭 - [RFC2144] Adams, C., "The CAST-128 Encryption Algorithm", RFC 2144, DOI 10.17487/RFC2144, May 1997,
<https://www.rfc-editor.org/info/rfc2144>. ⭐ - [RFC2822] Resnick, P., Ed., "Internet Message Format", RFC 2822, DOI 10.17487/RFC2822, April 2001,
<https://www.rfc-editor.org/info/rfc2822>. 🔥 - [RFC3156] Elkins, M., Del Torto, D., Levien, R., and T. Roessler, "MIME Security with OpenPGP", RFC 3156, DOI 10.17487/RFC3156, August 2001,
<https://www.rfc-editor.org/info/rfc3156>. 🌈 - [RFC3394] Schaad, J. and R. Housley, "Advanced Encryption Standard (AES) Key Wrap Algorithm", RFC 3394, DOI 10.17487/RFC3394, September 2002,
<https://www.rfc-editor.org/info/rfc3394>. 🎀 - [RFC3629] Yergeau, F., "UTF-8, a transformation format of ISO 10646", STD 63, RFC 3629, DOI 10.17487/RFC3629, November 2003,
<https://www.rfc-editor.org/info/rfc3629>. 🦋 - [RFC3713] Matsui, M., Nakajima, J., and S. Moriai, "A Description of the Camellia Encryption Algorithm", RFC 3713, DOI 10.17487/RFC3713, April 2004,
<https://www.rfc-editor.org/info/rfc3713>. 🌸 - [RFC4086] Eastlake 3rd, D., Schiller, J., and S. Crocker, "Randomness Requirements for Security", BCP 106, RFC 4086, DOI 10.17487/RFC4086, June 2005,
<https://www.rfc-editor.org/info/rfc4086>. 💜 - [RFC4648] Josefsson, S., "The Base16, Base32, and Base64 Data Encodings", RFC 4648, DOI 10.17487/RFC4648, October 2006,
<https://www.rfc-editor.org/info/rfc4648>. 🧚 - [RFC5322] Resnick, P., Ed., "Internet Message Format", RFC 5322, DOI 10.17487/RFC5322, October 2008,
<https://www.rfc-editor.org/info/rfc5322>. ⭐ - [RFC6234] Eastlake 3rd, D. and T. Hansen, "US Secure Hash Algorithms (SHA and SHA-based HMAC and HKDF)", RFC 6234, DOI 10.17487/RFC6234, May 2011,
<https://www.rfc-editor.org/info/rfc6234>. 🔥 - [RFC7253] Krovetz, T. and P. Rogaway, "The OCB Authenticated-Encryption Algorithm", RFC 7253, DOI 10.17487/RFC7253, May 2014,
<https://www.rfc-editor.org/info/rfc7253>. 🌈 - [RFC7748] Langley, A., Hamburg, M., and S. Turner, "Elliptic Curves for Security", RFC 7748, DOI 10.17487/RFC7748, January 2016,
<https://www.rfc-editor.org/info/rfc7748>. 🎀 - [RFC8017] Moriarty, K., Ed., Kaliski, B., Jonsson, J., and A. Rusch, "PKCS #1: RSA Cryptography Specifications Version 2.2", RFC 8017, DOI 10.17487/RFC8017, November 2016,
<https://www.rfc-editor.org/info/rfc8017>. 🦋 - [RFC8018] Moriarty, K., Ed., Kaliski, B., and A. Rusch, "PKCS #5: Password-Based Cryptography Specification Version 2.1", RFC 8018, DOI 10.17487/RFC8018, January 2017,
<https://www.rfc-editor.org/info/rfc8018>. 🌸 - [RFC8032] Josefsson, S. and I. Liusvaara, "Edwards-Curve Digital Signature Algorithm (EdDSA)", RFC 8032, DOI 10.17487/RFC8032, January 2017,
<https://www.rfc-editor.org/info/rfc8032>. 💜 - [RFC8126] Cotton, M., Leiba, B., and T. Narten, "Guidelines for Writing an IANA Considerations Section in RFCs", BCP 26, RFC 8126, DOI 10.17487/RFC8126, June 2017,
<https://www.rfc-editor.org/info/rfc8126>. 🧚 - [RFC8174] Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC 2119 Key Words", BCP 14, RFC 8174, DOI 10.17487/RFC8174, May 2017,
<https://www.rfc-editor.org/info/rfc8174>. ⭐ this one clears up that uppercase-vs-lowercase confusion, which is literally why the capitals matter 🙈 - [RFC9106] Biryukov, A., Dinu, D., Khovratovich, D., and S. Josefsson, "Argon2 Memory-Hard Function for Password Hashing and Proof-of-Work Applications", RFC 9106, DOI 10.17487/RFC9106, September 2021,
<https://www.rfc-editor.org/info/rfc9106>. 🔥 - [RIPEMD-160] ISO, "Information technology - Security techniques - Hash-functions - Part 3: Dedicated hash-functions", ISO/IEC 10118-3:1998, May 1998. 🌈
- [SP800-38A] NIST, "Recommendation for Block Cipher Modes of Operation: Methods and Techniques", NIST Special Publication 800-38A, DOI 10.6028/NIST.SP.800-38A, December 2001,
<https://nvlpubs.nist.gov/nistpubs/legacy/sp/nistspecialpublication800-38a.pdf>. 🎀 - [SP800-38D] NIST, "Recommendation for Block Cipher Modes of Operation: Galois/Counter Mode (GCM) and GMAC", NIST Special Publication 800-38D, DOI 10.6028/NIST.SP.800-38D, November 2007,
<https://nvlpubs.nist.gov/nistpubs/legacy/sp/nistspecialpublication800-38d.pdf>. 🦋 - [SP800-56A] NIST, "Recommendation for Pair-Wise Key Establishment Schemes Using Discrete Logarithm Cryptography", NIST Special Publication 800-56A Revision 3, DOI 10.6028/NIST.SP.800-56Ar, April 2018,
<https://nvlpubs.nist.gov/nistpubs/SpecialPublications/nist.sp.800-56Ar3.pdf>. 🌸 - [SP800-67] NIST, "Recommendation for the Triple Data Encryption Algorithm (TDEA) Block Cipher", NIST Special Publication 800-67 Revision 2, DOI 10.6028/NIST.SP.800-67r2, November 2017,
<https://nvlpubs.nist.gov/nistpubs/SpecialPublications/NIST.SP.800-67r2.pdf>. 💜 - [TWOFISH] Schneier, B., Kelsey, J., Whiting, D., Wagner, D., Hall, C., and N. Ferguson, "Twofish: A 128-Bit Block Cipher", June 1998,
<https://www.schneier.com/wp-content/uploads/2016/02/paper-twofish-paper.pdf>. 🧚
16.2. Informative References 📚✨
These are the extra reads that help you understand the spec, bestie 🧠💖
[BLEICHENBACHER]Bleichenbacher, D., "Generating ElGamal Signatures Without Knowing the Secret Key", EUROCRYPT'96: International Conference on the Theory and Applications of Cryptographic Techniques Proceedings, Vol. 1070, pp. 10-18, May 1996.[BLEICHENBACHER-PKCS1]Bleichenbacher, D., "Chosen Ciphertext Attacks Against Protocols Based on the RSA Encryption Standard PKCS #1", CRYPTO '98: International Cryptology Conference Proceedings, Vol. 1462, pp. 1-12, August 1998,<http://archiv.infsec.ethz.ch/education/fs08/secsem/Bleichenbacher98.pdf>.[C99]ISO, "Information technology - Programming languages: C", ISO/IEC 9899:2018, June 2018,<https://www.iso.org/standard/74528.html>.[CHECKOWAY]Checkoway, S., Maskiewicz, J., Garman, C., Fried, J., Cohney, S., Green, M., Heninger, N., Weinmann, RP., Rescorla, E., and H. Shacham, "A Systematic Analysis of the Juniper Dual EC Incident", Proceedings of the 2016 ACM SIGSAC Conference on Computer and Communications Security, DOI 10.1145/2976749.2978395, October 2016,<https://doi.org/10.1145/2976749.2978395>.[EFAIL]Poddebniak, D., Dresen, C., Müller, J., Ising, F., Schinzel, S., Friedberger, S., Somorovsky, J., and J. Schwenk, "Efail: Breaking S/MIME and OpenPGP Email Encryption using Exfiltration Channels", Proceedings of the 27th USENIX Security Symposium, August 2018,<https://www.usenix.org/system/files/conference/usenixsecurity18/sec18-poddebniak.pdf>.[Errata-2199]RFC Errata, Erratum ID 2199, RFC 4880,<https://www.rfc-editor.org/errata/eid2199>.[Errata-2200]RFC Errata, Erratum ID 2200, RFC 4880,<https://www.rfc-editor.org/errata/eid2200>.[Errata-2206]RFC Errata, Erratum ID 2206, RFC 4880,<https://www.rfc-editor.org/errata/eid2206>.[Errata-2208]RFC Errata, Erratum ID 2208, RFC 4880,<https://www.rfc-editor.org/errata/eid2208>.[Errata-2214]RFC Errata, Erratum ID 2214, RFC 4880,<https://www.rfc-editor.org/errata/eid2214>.[Errata-2216]RFC Errata, Erratum ID 2216, RFC 4880,<https://www.rfc-editor.org/errata/eid2216>.[Errata-2219]RFC Errata, Erratum ID 2219, RFC 4880,<https://www.rfc-editor.org/errata/eid2219>.[Errata-2222]RFC Errata, Erratum ID 2222, RFC 4880,<https://www.rfc-editor.org/errata/eid2222>.[Errata-2226]RFC Errata, Erratum ID 2226, RFC 4880,<https://www.rfc-editor.org/errata/eid2226>.[Errata-2234]RFC Errata, Erratum ID 2234, RFC 4880,<https://www.rfc-editor.org/errata/eid2234>.[Errata-2235]RFC Errata, Erratum ID 2235, RFC 4880,<https://www.rfc-editor.org/errata/eid2235>.[Errata-2236]RFC Errata, Erratum ID 2236, RFC 4880,<https://www.rfc-editor.org/errata/eid2236>.[Errata-2238]RFC Errata, Erratum ID 2238, RFC 4880,<https://www.rfc-editor.org/errata/eid2238>.[Errata-2240]RFC Errata, Erratum ID 2240, RFC 4880,<https://www.rfc-editor.org/errata/eid2240>.[Errata-2242]RFC Errata, Erratum ID 2242, RFC 4880,<https://www.rfc-editor.org/errata/eid2242>.[Errata-2243]RFC Errata, Erratum ID 2243, RFC 4880,<https://www.rfc-editor.org/errata/eid2243>.[Errata-2270]RFC Errata, Erratum ID 2270, RFC 4880,<https://www.rfc-editor.org/errata/eid2270>.[Errata-2271]RFC Errata, Erratum ID 2271, RFC 4880,<https://www.rfc-editor.org/errata/eid2271>.[Errata-3298]RFC Errata, Erratum ID 3298, RFC 4880,<https://www.rfc-editor.org/errata/eid3298>.[Errata-5491]RFC Errata, Erratum ID 5491, RFC 4880,<https://www.rfc-editor.org/errata/eid5491>.[Errata-7545]RFC Errata, Erratum ID 7545, RFC 4880,<https://www.rfc-editor.org/errata/eid7545>.[Errata-7889]RFC Errata, Erratum ID 7889, RFC 4880,<https://www.rfc-editor.org/errata/eid7889>.[HASTAD]Hastad, J., "Solving Simultaneous Modular Equations of Low Degree", DOI 10.1137/0217019, April 1988,<https://doi.org/10.1137/0217019>.[JKS02]Jallad, K., Katz, J., and B. Schneier, "Implementation of Chosen-Ciphertext Attacks against PGP and GnuPG", DOI 0.1007/3-540-45811-5_7, September 2002,<https://www.schneier.com/academic/archives/2002/01/implementation_of_ch.html>.[KOBLITZ]Koblitz, N., "A course in number theory and cryptography", Chapter VI: Elliptic Curves, DOI 10.2307/3618498, 1997,<https://doi.org/10.2307/3618498>.[KOPENPGP]Bruseghini, L., Paterson, K. G., and D. Huigens, "Victory by KO: Attacking OpenPGP Using Key Overwriting", Proceedings of the ACM SIGSAC Conference on Computer and Communications Security, pp. 411-423, DOI 10.1145/3548606.3559363, November 2022,<https://dl.acm.org/doi/10.1145/3548606.3559363>.[KR02]Klíma, V. and T. Rosa, "Attack on Private Signature Keys of the OpenPGP Format, PGP(TM) Programs and Other Applications Compatible with OpenPGP", Cryptology ePrint Archive, Paper 2002/076, March 2001,<https://eprint.iacr.org/2002/076>.[MRLG15]Maury, F., Reinhard, JR., Levillain, O., and H. Gilbert, "Format Oracles on OpenPGP", Topics in Cryptology -- CT-RSA 2015, Vol. 9048, pp. 220-236, DOI 10.1007/978-3-319-16715-2_12, January 2015,<https://doi.org/10.1007/978-3-319-16715-2_12>.[MZ05]Mister, S. and R. Zuccherato, "An Attack on CFB Mode Encryption As Used By OpenPGP", Cryptology ePrint Archive, Paper 2005/033, February 2005,<http://eprint.iacr.org/2005/033>.[OPENPGPCARD]Pietig, A., "Functional Specification of the OpenPGP application on ISO Smart Card Operating Systems", Version 3.4.1, March 2020,<https://gnupg.org/ftp/specs/OpenPGP-smart-card-application-3.4.1.pdf>.[PAX]The Open Group, "The Open Group Base Specifications", 'pax - portable archive interchange', Issue 7, 2018 Edition, IEEE Std 1003.1-2017, 2018,<https://pubs.opengroup.org/onlinepubs/9699919799/utilities/pax.html>.[PSSLR17]Poddebniak, D., Somorovsky, J., Schinzel, S., Lochter, M., and P. Rösler, "Attacking Deterministic Signature Schemes using Fault Attacks", Cryptology ePrint Archive, Paper 2017/1014, October 2017,<https://eprint.iacr.org/2017/1014>.[REGEX]regex, "Henry Spencer's regular expression libraries",<https://garyhouston.github.io/regex/>.[RFC1991]Atkins, D., Stallings, W., and P. Zimmermann, "PGP Message Exchange Formats", RFC 1991, DOI 10.17487/RFC1991, August 1996,<https://www.rfc-editor.org/info/rfc1991>.[RFC2440]Callas, J., Donnerhacke, L., Finney, H., and R. Thayer, "OpenPGP Message Format", RFC 2440, DOI 10.17487/RFC2440, November 1998,<https://www.rfc-editor.org/info/rfc2440>.[RFC2978]Freed, N. and J. Postel, "IANA Charset Registration Procedures", BCP 19, RFC 2978, DOI 10.17487/RFC2978, October 2000,<https://www.rfc-editor.org/info/rfc2978>.[RFC4880]Callas, J., Donnerhacke, L., Finney, H., Shaw, D., and R. Thayer, "OpenPGP Message Format", RFC 4880, DOI 10.17487/RFC4880, November 2007,<https://www.rfc-editor.org/info/rfc4880>.[RFC5581]Shaw, D., "The Camellia Cipher in OpenPGP", RFC 5581, DOI 10.17487/RFC5581, June 2009,<https://www.rfc-editor.org/info/rfc5581>.[RFC5639]Lochter, M. and J. Merkle, "Elliptic Curve Cryptography (ECC) Brainpool Standard Curves and Curve Generation", RFC 5639, DOI 10.17487/RFC5639, March 2010,<https://www.rfc-editor.org/info/rfc5639>.[RFC5869]Krawczyk, H. and P. Eronen, "HMAC-based Extract-and-Expand Key Derivation Function (HKDF)", RFC 5869, DOI 10.17487/RFC5869, May 2010,<https://www.rfc-editor.org/info/rfc5869>.[RFC6090]McGrew, D., Igoe, K., and M. Salter, "Fundamental Elliptic Curve Cryptography Algorithms", RFC 6090, DOI 10.17487/RFC6090, February 2011,<https://www.rfc-editor.org/info/rfc6090>.[RFC6637]Jivsov, A., "Elliptic Curve Cryptography (ECC) in OpenPGP", RFC 6637, DOI 10.17487/RFC6637, June 2012,<https://www.rfc-editor.org/info/rfc6637>.[SEC1]Standards for Efficient Cryptography Group, "SEC 1: Elliptic Curve Cryptography", May 2009,<https://www.secg.org/sec1-v2.pdf>.[SHA1CD]"sha1collisiondetection", commit b4a7b0b, December 2020,<https://github.com/cr-marcstevens/sha1collisiondetection>.[SHAMBLES]Leurent, G. and T. Peyrin, "Sha-1 is a shambles: first chosen-prefix collision on sha-1 and application to the PGP web of trust", August 2020,<https://dl.acm.org/doi/abs/10.5555/3489212.3489316/>.[SP800-131A]NIST, "Transitioning the Use of Cryptographic Algorithms and Key Lengths", NIST Special Publication 800-131A, Revision 2, DOI 10.6028/NIST.SP.800-131Ar2, March 2019,<https://nvlpubs.nist.gov/nistpubs/SpecialPublications/NIST.SP.800-131Ar2.pdf>.[SP800-57]NIST, "Recommendation for Key Management: Part 1 - General", NIST Special Publication 800-57 Part 1, Revision 5, DOI 10.6028/NIST.SP.800-57pt1r5, May 2020,<https://nvlpubs.nist.gov/nistpubs/SpecialPublications/NIST.SP.800-57pt1r5.pdf>.[STEVENS2013]Stevens, M., "Counter-cryptanalysis", Cryptology ePrint Archive, Paper 2013/358, June 2013,<https://eprint.iacr.org/2013/358>.[UNIFIED-DIFF]Free Software Foundation, "Comparing and Merging Files", 'Detailed Description of Unified Format', Section 2.2.2.2, January 2021,<https://www.gnu.org/software/diffutils/manual/html_node/Detailed-Unified.html>.[USENIX-STUDY]Dechand, S., Schürmann, D., Busse, K., Acar, Y., Fahl, S., and M. Smith, "An Empirical Study of Textual Key-Fingerprint Representations", ISBN 978-1-931971-32-4, August 2016,<https://www.usenix.org/system/files/conference/usenixsecurity16/sec16_paper_dechand.pdf>.
Appendix A. Test Vectors 🧪✨
Okay so this part is basically the "here are some examples you can check your code against" section 💅 To help with the implementation of this specification, a set of non-normative examples follow. Non-normative means they're helpful but they aren't the rules themselves, so don't treat them as the spec 🫶
A.1. Sample Version 4 Ed25519Legacy Key 🔑
okay bestie, here's the secret key we're using for this example 🗝️
D: 1a8b1ff05ded48e18bf50166c664ab023ea70003d78d9e41f5758a91d850f8d2
fr, this is the raw secret key that goes into the EdDSA signing operation ✨ The key was created on 2014-08-19 14:28:27, so the fingerprint of the OpenPGP Key is:
C959 BDBA FA32 A2F8 9A15 3B67 8CFD E121 9796 5A9AThe algorithm-specific input parameters, without the MPI length headers, are:
oid: 2b06010401da470f01
q: 403f098994bdd916ed4053197934e4a87c80733a1280d62f8010992e43ee3b2406
The whole Public Key packet looks like this 🌸
98 33 04 53 f3 5f 0b 16 09 2b 06 01 04 01 da 47
0f 01 01 07 40 3f 09 89 94 bd d9 16 ed 40 53 19
79 34 e4 a8 7c 80 73 3a 12 80 d6 2f 80 10 99 2e
43 ee 3b 24 06And here's the same packet in ASCII-armored form, which is giving very copy-paste energy 💅
-----BEGIN PGP PUBLIC KEY BLOCK-----
xjMEU/NfCxYJKwYBBAHaRw8BAQdAPwmJlL3ZFu1AUxl5NOSofIBzOhKA1i+AEJku
Q+47JAY=
-----END PGP PUBLIC KEY BLOCK-----A.2. Sample Version 4 Ed25519Legacy Signature ✍️
So we sign the input data "OpenPGP" with the sample key on 2015-09-16 12:24:53 UTC 🥰 The input to the hash function ends up being:
m: 4f70656e504750040016080006050255f95f9504ff0000000c
Using the SHA2-256 hash algorithm gives us this digest:
d: f6220a3f757814f4c2176ffbb68b00249cd4ccdc059c4b34ad871f30b1740280
That digest gets fed into the EdDSA signature function, and it spits out this signature:
r: 56f90cca98e2102637bd983fdb16c131dfd27ed82bf4dde5606e0d756aed3366
s: d09c4fa11527f038e0f57f2201d82f2ea2c9033265fa6ceb489e854bae61b404
The whole Signature packet is:
88 5e 04 00 16 08 00 06 05 02 55 f9 5f 95 00 0a
09 10 8c fd e1 21 97 96 5a 9a f6 22 00 ff 56 f9
0c ca 98 e2 10 26 37 bd 98 3f db 16 c1 31 df d2
7e d8 2b f4 dd e5 60 6e 0d 75 6a ed 33 66 01 00
d0 9c 4f a1 15 27 f0 38 e0 f5 7f 22 01 d8 2f 2e
a2 c9 03 32 65 fa 6c eb 48 9e 85 4b ae 61 b4 04Same packet, ASCII-armored 🫶
-----BEGIN PGP SIGNATURE-----
iF4EABYIAAYFAlX5X5UACgkQjP3hIZeWWpr2IgD/VvkMypjiECY3vZg/2xbBMd/S
ftgr9N3lYG4NdWrtM2YBANCcT6EVJ/A44PV/IgHYLy6iyQMyZfps60iehUuuYbQE
-----END PGP SIGNATURE-----A.3. Sample Version 6 Certificate (Transferable Public Key) 👑
Okay this one is the main character 🦋 It's a Transferable Public Key made of:
- A version 6 Ed25519 Public Key packet
- A version 6 Direct Key self-signature
- A version 6 X25519 Public Subkey packet
- A version 6 Subkey Binding signature
The primary key has this fingerprint:
CB186C4F0609A697E4D52DFA6C722B0C1F1E27C18A56708F6525EC27BAD9ACC9The subkey has this fingerprint:
12C83F1E706F6308FE151A417743A1F033790E93E9978488D1DB378DA9930885And the certificate itself:
-----BEGIN PGP PUBLIC KEY BLOCK-----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-----END PGP PUBLIC KEY BLOCK-----The matching Transferable Secret Key is in Appendix A.4, so go check that out 👀
A.3.1. Hashed Data Stream for Signature Verification 🧩
Okay so the Direct Key self-signature in the certificate from Appendix A.3 is made over this sequence of data:
0x0000 10 3e 2d 7d 22 7e c0 e6
0x0008 d7 ce 44 71 db 36 bf c9
0x0010 70 83 25 36 90 27 14 98
0x0018 a7 ef 05 76 c0 7f aa e1
0x0020 9b 00 00 00 2a 06 63 87
0x0028 7f e3 1b 00 00 00 20 f9
0x0030 4d a7 bb 48 d6 0a 61 e5
0x0038 67 70 6a 65 87 d0 33 19
0x0040 99 bb 9d 89 1a 08 24 2e
0x0048 ad 84 54 3d f8 95 a3 06
0x0050 1f 1b 0a 00 00 00 42 05
0x0058 82 63 87 7f e3 03 0b 09
0x0060 07 05 15 0a 0e 08 0c 02
0x0068 16 00 02 9b 03 02 1e 09
0x0070 22 21 06 cb 18 6c 4f 06
0x0078 09 a6 97 e4 d5 2d fa 6c
0x0080 72 2b 0c 1f 1e 27 c1 8a
0x0088 56 70 8f 65 25 ec 27 ba
0x0090 d9 ac c9 05 27 09 02 07
0x0098 02 06 ff 00 00 00 4aSame data, but now broken out by octet and what each one means 🤓
0x0000 10 3e 2d 7d 22 7e c0 e6 salt
0x0008 d7 ce 44 71 db 36 bf c9
0x0010 70 83 25 36 90 27 14 98
0x0018 a7 ef 05 76 c0 7f aa e1
[ pubkey begins ]
0x0020 9b key packet
0x0021 00 00 00 2a pubkey length
0x0025 06 pubkey version
0x0026 63 87 creation time
0x0028 7f e3 (2022-11-30T16:08:03Z)
0x002a 1b key algo: Ed25519
0x002b 00 00 00 20 key length
0x002f f9 Ed25519 public key
0x0030 4d a7 bb 48 d6 0a 61 e5
0x0038 67 70 6a 65 87 d0 33 19
0x0040 99 bb 9d 89 1a 08 24 2e
0x0048 ad 84 54 3d f8 95 a3
[ trailer begins ]
0x004f 06 sig version 6
0x0050 1f sig type: Direct Key signature
0x0051 1b sig algo: Ed25519
0x0052 0a hash ago: SHA2-512
0x0053 00 00 00 42 hashed subpackets length
0x0057 05 subpkt length
0x0058 82 critical subpkt: Sig Creation Time
0x0059 63 87 7f e3 Signature Creation Time
0x005d 03 subpkt length
0x005e 0b subpkt type: Pref. v1 SEIPD Ciphers
0x005f 09 Ciphers: [AES256 AES128]
0x0060 07
0x0061 05 subpkt length
0x0062 15 subpkt type: Pref. Hash Algorithms
0x0063 0a 0e Hashes: [SHA2-512 SHA3-512
0x0065 08 0c SHA2-256 SHA3-256]
0x0067 02 subpkt length
0x0068 16 subpkt type: Pref. Compression
0x0069 00 Compression: [none]
0x006a 02 subpkt length
0x006b 9b critical subpkt: Key Flags
0x006c 03 Key Flags: {certify, sign}
0x006d 02 subpkt length
0x006e 1e subpkt type: Features
0x006f 09 Features: {v1SEIPD, v2SEIPD}
0x0070 22 subpkt length
0x0071 21 subpkt type: Issuer Fingerprint
0x0072 06 Fingerprint version 6
0x0073 cb 18 6c 4f 06 Fingerprint
0x0078 09 a6 97 e4 d5 2d fa 6c
0x0080 72 2b 0c 1f 1e 27 c1 8a
0x0088 56 70 8f 65 25 ec 27 ba
0x0090 d9 ac c9
0x0093 05 subpkt length
0x0094 27 subpkt type: Pref. AEAD Ciphersuites
0x0095 09 02 07 Ciphersuites:
0x0098 02 [ AES256-OCB, AES128-OCB ]
0x0099 06 sig version 6
0x009a ff sentinel octet
0x009b 00 00 00 4a trailer lengthThe Subkey Binding signature from Appendix A.3 is made over this sequence of data ✨
0x0000 a6 e9 18 6d 9d 59 35 fc
0x0008 8f e5 63 14 cd b5 27 48
0x0010 6a 5a 51 20 f9 b7 62 a2
0x0018 35 a7 29 f0 39 01 0a 56
0x0020 9b 00 00 00 2a 06 63 87
0x0028 7f e3 1b 00 00 00 20 f9
0x0030 4d a7 bb 48 d6 0a 61 e5
0x0038 67 70 6a 65 87 d0 33 19
0x0040 99 bb 9d 89 1a 08 24 2e
0x0048 ad 84 54 3d f8 95 a3 9b
0x0050 00 00 00 2a 06 63 87 7f
0x0058 e3 19 00 00 00 20 86 93
0x0060 24 83 67 f9 e5 01 5d b9
0x0068 22 f8 f4 80 95 dd a7 84
0x0070 98 7f 2d 59 85 b1 2f ba
0x0078 d1 6c af 5e 44 35 06 18
0x0080 1b 0a 00 00 00 2c 05 82
0x0088 63 87 7f e3 02 9b 0c 22
0x0090 21 06 cb 18 6c 4f 06 09
0x0098 a6 97 e4 d5 2d fa 6c 72
0x00a0 2b 0c 1f 1e 27 c1 8a 56
0x00a8 70 8f 65 25 ec 27 ba d9
0x00b0 ac c9 06 ff 00 00 00 34And the same data, broken out by octet and semantics, again because the details matter a LOT here 💖
0x0000 a6 e9 18 6d 9d 59 35 fc salt
0x0008 8f e5 63 14 cd b5 27 48
0x0010 6a 5a 51 20 f9 b7 62 a2
0x0018 35 a7 29 f0 39 01 0a 56
[ primary pubkey begins ]
0x0020 9b key packet
0x0021 00 00 00 2a pubkey length
0x0025 06 pubkey version
0x0026 63 87 creation time
0x0028 7f e3 (2022-11-30T16:08:03Z)
0x002a 1b key algo: Ed25519
0x002b 00 00 00 20 key length
0x002f f9 Ed25519 public key
0x0030 4d a7 bb 48 d6 0a 61 e5
0x0038 67 70 6a 65 87 d0 33 19
0x0040 99 bb 9d 89 1a 08 24 2e
0x0048 ad 84 54 3d f8 95 a3
[ subkey pubkey begins ]
0x004f 9b key packet
0x0050 00 00 00 2a pubkey length
0x0054 06 pubkey version
0x0055 63 87 7f creation time (2022-11-30T16:08:03Z)
0x0058 e3
0x0059 19 key algo: X25519
0x005a 00 00 00 20 key length
0x005e 86 93 X25519 public key
0x0060 24 83 67 f9 e5 01 5d b9
0x0068 22 f8 f4 80 95 dd a7 84
0x0070 98 7f 2d 59 85 b1 2f ba
0x0078 d1 6c af 5e 44 35
[ trailer begins ]
0x007e 06 sig version 6
0x007f 18 sig type: Subkey Binding sig
0x0080 1b sig algo Ed25519
0x0081 0a hash algo: SHA2-512
0x0082 00 00 00 2c hashed subpackets length
0x0086 05 subpkt length
0x0087 82 critical subpkt: Sig Creation Time
0x0088 63 87 7f e3 Signature Creation Time
0x008c 02 subpkt length
0x008d 9b critical subpkt: Key Flags
0x008e 0c Key Flags: {EncComms, EncStorage}
0x008f 22 subpkt length
0x0090 21 subpkt type: Issuer Fingerprint
0x0091 06 Fingerprint version 6
0x0092 cb 18 6c 4f 06 09 Fingerprint
0x0098 a6 97 e4 d5 2d fa 6c 72
0x00a0 2b 0c 1f 1e 27 c1 8a 56
0x00a8 70 8f 65 25 ec 27 ba d9
0x00b0 ac c9
0x00b2 06 sig version 6
0x00b3 ff sentinel octet
0x00b4 00 00 00 34 trailer lengthA.4. Sample Version 6 Secret Key (Transferable Secret Key) 🔐✨
okay bestie, this is the secret key bundle! 🗝️💖 It's a Transferable Secret Key made of four pieces:
- A version 6 Ed25519 Secret Key packet 🦋
- A version 6 Direct Key self-signature 💅
- A version 6 X25519 Secret Subkey packet 🌸
- A version 6 Subkey Binding signature 🫶
-----BEGIN PGP PRIVATE KEY BLOCK-----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-----END PGP PRIVATE KEY BLOCK-----The matching Transferable Public Key is over in Appendix A.3 🎀, so go check it out!
A.5. Sample Locked Version 6 Secret Key (Transferable Secret Key) 🔒💜
this is the SAME secret key as A.4, just locked up tight 🔐 with a passphrase, using AEAD and Argon2. so secure, periodt 💅
The passphrase is the ASCII string:
correct horse battery staple 🐴🔋
-----BEGIN PGP PRIVATE KEY BLOCK-----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-----END PGP PRIVATE KEY BLOCK-----A.5.1. Intermediate Data for Locked Primary Key 🧁
the S2K-derived material for the primary key is:
832bd2662a5c2b251ee3fc82aec349a766ca539015880133002e5a21960b3bcfafter HKDF, the symmetric key used for AEAD encryption of the primary key is:
9e37cb26787f37e18db172795c4c297550d39ac82511d9af4c8706db6a77fd51the additional data for AEAD for the primary key is:
c50663877fe31b00000020f94da7bb48d60a61e567706a6587d0331999bb9d89
1a08242ead84543df895a3A.5.2. Intermediate Data for Locked Subkey 🍓
the S2K-derived key material for the subkey is:
f74a6ce873a089ef13a3da9ac059777bb22340d15eaa6c9dc0f8ef09035c67cdafter HKDF, the symmetric key used for AEAD encryption of the subkey is:
3c60cb63285f62f4c3de49835786f011cf6f4c069f61232cd7013ff5fd31e603the additional data for AEAD for the subkey is:
c70663877fe319000000208693248367f9e5015db922f8f48095dda784987f2d
5985b12fbad16caf5e4435A.6. Sample Cleartext Signed Message 📝🌈
okay so here's a signed message using the Cleartext Signature Framework (Section 7) 🫧. You can verify it with the certificate from Appendix A.3, no cap.
Heads up, this message uses dash-escaping (Section 7.2) because of what's inside it 🙈.
-----BEGIN PGP SIGNED MESSAGE-----
What we need from the grocery store:
- - tofu
- - vegetables
- - noodles
-----BEGIN PGP SIGNATURE-----
wpgGARsKAAAAKQWCY5ijYyIhBssYbE8GCaaX5NUt+mxyKwwfHifBilZwj2Ul7Ce6
2azJAAAAAGk2IHZJX1AhiJD39eLuPBgiUU9wUA9VHYblySHkBONKU/usJ9BvuAqo
/FvLFuGWMbKAdA+epq7V4HOtAPlBWmU8QOd6aud+aSunHQaaEJ+iTFjP2OMW0KBr
NK2ay45cX1IVAQ==
-----END PGP SIGNATURE-----the Signature packet here is:
0x0000 c2 packet type: Signature
0x0001 98 packet length
0x0002 06 sig version 6
0x0003 01 sig type: Canonical Text
0x0004 1b pubkey algorithm: Ed25519
0x0005 0a hash algorithm used: SHA2-512
0x0006 00 00 hashed subpackets length: 41
0x0008 00 29
0x000a 05 subpkt length
0x000b 82 critical subpkt: Sig Creation Time
0x000c 63 98 a3 63 (2022-12-13T16:08:03Z)
0x0010 22 subpkt length
0x0011 21 subpkt type: Issuer Fingerprint
0x0012 06 Fingerprint version 6
0x0013 cb 18 6c 4f 06 Fingerprint
0x001a 09 a6 97 e4 d5 2d fa 6c
0x0020 72 2b 0c 1f 1e 27 c1 8a
0x0028 56 70 8f 65 25 ec 27 ba
0x0030 d9 ac c9
0x0033 00 00 00 00 unhashed subpackets length: 0
0x0037 69 left 16 bits of signed hash
0x0038 36
0x0039 20 salt length
0x003a 76 49 5f 50 21 88 salt
0x0040 90 f7 f5 e2 ee 3c 18 22
0x0048 51 4f 70 50 0f 55 1d 86
0x0050 e5 c9 21 e4 04 e3 4a 53
0x0058 fb ac
0x005a 27 d0 6f b8 0a a8 Ed25519 signature
0x0060 fc 5b cb 16 e1 96 31 b2
0x0068 80 74 0f 9e a6 ae d5 e0
0x0070 73 ad 00 f9 41 5a 65 3c
0x0078 40 e7 7a 6a e7 7e 69 2b
0x0080 a7 1d 06 9a 10 9f a2 4c
0x0088 58 cf d8 e3 16 d0 a0 6b
0x0090 34 ad 9a cb 8e 5c 5f 52
0x0098 15 01the signature is made over the following data:
0x0000 76 49 5f 50 21 88 90 f7
0x0008 f5 e2 ee 3c 18 22 51 4f
0x0010 70 50 0f 55 1d 86 e5 c9
0x0018 21 e4 04 e3 4a 53 fb ac
0x0020 57 68 61 74 20 77 65 20
0x0028 6e 65 65 64 20 66 72 6f
0x0030 6d 20 74 68 65 20 67 72
0x0038 6f 63 65 72 79 20 73 74
0x0040 6f 72 65 3a 0d 0a 0d 0a
0x0048 2d 20 74 6f 66 75 0d 0a
0x0050 2d 20 76 65 67 65 74 61
0x0058 62 6c 65 73 0d 0a 2d 20
0x0060 6e 6f 6f 64 6c 65 73 0d
0x0068 0a 06 01 1b 0a 00 00 00
0x0070 29 05 82 63 98 a3 63 22
0x0078 21 06 cb 18 6c 4f 06 09
0x0080 a6 97 e4 d5 2d fa 6c 72
0x0088 2b 0c 1f 1e 27 c1 8a 56
0x0090 70 8f 65 25 ec 27 ba d9
0x0098 ac c9 06 ff 00 00 00 31the same data, broken out by octet and semantics, is:
0x0000 76 49 5f 50 21 88 90 f7 salt
0x0008 f5 e2 ee 3c 18 22 51 4f
0x0010 70 50 0f 55 1d 86 e5 c9
0x0018 21 e4 04 e3 4a 53 fb ac
[ message begins ]
0x0020 57 68 61 74 20 77 65 20 canonicalized message
0x0028 6e 65 65 64 20 66 72 6f
0x0030 6d 20 74 68 65 20 67 72
0x0038 6f 63 65 72 79 20 73 74
0x0040 6f 72 65 3a 0d 0a 0d 0a
0x0048 2d 20 74 6f 66 75 0d 0a
0x0050 2d 20 76 65 67 65 74 61
0x0058 62 6c 65 73 0d 0a 2d 20
0x0060 6e 6f 6f 64 6c 65 73 0d
0x0068 0a
[ trailer begins ]
0x0069 06 sig version 6
0x006a 01 sig type: Canonical Text
0x006b 1b pubkey algorithm: Ed25519
0x006c 0a hash algorithm: SHA2-512
0x006d 00 00 00 hashed subpackets length
0x0070 29
0x0071 05 subpacket length
0x0072 82 critical subpkt: Sig Creation Time
0x0073 63 98 a3 63 (2022-12-13T16:08:03Z)
0x0077 22 subpkt length
0x0078 21 subpkt type: Issuer Fingerprint
0x0079 06 Fingerprint version 6
0x007a cb 18 6c 4f 06 09 Fingerprint
0x0080 a6 97 e4 d5 2d fa 6c 72
0x0088 2b 0c 1f 1e 27 c1 8a 56
0x0090 70 8f 65 25 ec 27 ba d9
0x0098 ac c9
0x009a 06 sig version 6
0x009b ff sentinel octet
0x009c 00 00 00 31 trailer lengththe calculated SHA2-512 hash digest over this data is:
69365bf44a97af1f0844f1f6ab83fdf6b36f26692efaa621a8aac91c4e29ea07
e894cabc6e2f20eedfce6c03b89141a2cc7cbe245e6e7a5654addbec5000b89bA.7. Sample Inline-Signed Message ✨📬
this is the same message and signature as A.6, but as an inline-signed message 💌. The hashed data is exactly the same, and all the intermediate values and annotated hex dumps still apply, iykyk 🦋.
-----BEGIN PGP MESSAGE-----
xEYGAQobIHZJX1AhiJD39eLuPBgiUU9wUA9VHYblySHkBONKU/usyxhsTwYJppfk
1S36bHIrDB8eJ8GKVnCPZSXsJ7rZrMkBy0p1AAAAAABXaGF0IHdlIG5lZWQgZnJv
bSB0aGUgZ3JvY2VyeSBzdG9yZToKCi0gdG9mdQotIHZlZ2V0YWJsZXMKLSBub29k
bGVzCsKYBgEbCgAAACkFgmOYo2MiIQbLGGxPBgmml+TVLfpscisMHx4nwYpWcI9l
JewnutmsyQAAAABpNiB2SV9QIYiQ9/Xi7jwYIlFPcFAPVR2G5ckh5ATjSlP7rCfQ
b7gKqPxbyxbhljGygHQPnqau1eBzrQD5QVplPEDnemrnfmkrpx0GmhCfokxYz9jj
FtCgazStmsuOXF9SFQE=
-----END PGP MESSAGE-----A.8. Sample X25519-AEAD-OCB Encryption and Decryption 🦄💜
this example encrypts the cleartext string Hello, world! for the sample cert (see Appendix A.3) 💌, using AES-128 with AEAD-OCB encryption. Let's gooo 🚀
A.8.1. Sample Version 6 Public Key Encrypted Session Key Packet 🎀
this packet contains the following series of octets:
0x0000 c1 5d 06 21 06 12 c8 3f
0x0008 1e 70 6f 63 08 fe 15 1a
0x0010 41 77 43 a1 f0 33 79 0e
0x0018 93 e9 97 84 88 d1 db 37
0x0020 8d a9 93 08 85 19 87 cf
0x0028 18 d5 f1 b5 3f 81 7c ce
0x0030 5a 00 4c f3 93 cc 89 58
0x0038 bd dc 06 5f 25 f8 4a f5
0x0040 09 b1 7d d3 67 64 18 de
0x0048 a3 55 43 79 56 61 79 01
0x0050 e0 69 57 fb ca 8a 6a 47
0x0058 a5 b5 15 3e 8d 3a b7the same data, broken out by octet and semantics, is:
0x0000 c1 packet type: PKESK
0x0001 5d packet length
0x0002 06 v6 PKESK
0x0003 21 length of fingerprint
0x0004 06 Key version 6
0x0005 12 c8 3f Key fingerprint
0x0008 1e 70 6f 63 08 fe 15 1a
0x0010 41 77 43 a1 f0 33 79 0e
0x0018 93 e9 97 84 88 d1 db 37
0x0020 8d a9 93 08 85
0x0025 19 algorithm: X25519
0x0026 87 cf Ephemeral key
0x0028 18 d5 f1 b5 3f 81 7c ce
0x0030 5a 00 4c f3 93 cc 89 58
0x0038 bd dc 06 5f 25 f8 4a f5
0x0040 09 b1 7d d3 67 64
0x0046 18 ESK length
0x0047 de ESK
0x0048 a3 55 43 79 56 61 79 01
0x0050 e0 69 57 fb ca 8a 6a 47
0x0058 a5 b5 15 3e 8d 3a b7A.8.2. X25519 Encryption/Decryption of the Session Key 🔑
ephemeral key:
87 cf 18 d5 f1 b5 3f 81 7c ce 5a 00 4c f3 93 cc
89 58 bd dc 06 5f 25 f8 4a f5 09 b1 7d d3 67 64this ephemeral key is derived from the following ephemeral secret key material, which is never placed on the wire 🙈:
af 1e 43 c0 d1 23 ef e8 93 a7 d4 d3 90 f3 a7 61
e3 fa c3 3d fc 7f 3e da a8 30 c9 01 13 52 c7 79public key from the target certificate (see Appendix A.3):
86 93 24 83 67 f9 e5 01 5d b9 22 f8 f4 80 95 dd
a7 84 98 7f 2d 59 85 b1 2f ba d1 6c af 5e 44 35the corresponding long-lived X25519 private key material (see Appendix A.4):
4d 60 0a 4f 79 4d 44 77 5c 57 a2 6e 0f ee fe d5
58 e9 af ff d6 ad 0d 58 2d 57 fb 2b a2 dc ed b8shared point:
67 e3 0e 69 cd c7 ba b2 a2 68 0d 78 ac a4 6a 2f
8b 6e 2a e4 4d 39 8b dc 6f 92 c5 ad 4a 49 25 14HKDF output:
f6 6d ad cf f6 45 92 23 9b 25 45 39 b6 4f f6 07decrypted session key:
dd 70 8f 6f a1 ed 65 11 4d 68 d2 34 3e 7c 2f 1dA.8.3. Sample v2 SEIPD Packet ✨
Okay, here's the full packet, bestie 🎀 Every byte is listed below in hex:
0x0000 d2 69 02 07 02 06 61 64
0x0008 16 53 5b e0 b0 71 6d 60
0x0010 e0 52 a5 6c 4c 40 7f 9e
0x0018 b3 6b 0e fa fe 9a d0 a0
0x0020 df 9b 03 3c 69 a2 1b a9
0x0028 eb d2 c0 ec 95 bf 56 9d
0x0030 25 c9 99 ee 4a 3d e1 70
0x0038 58 f4 0d fa 8b 4c 68 2b
0x0040 e3 fb bb d7 b2 7e b0 f5
0x0048 9b b5 00 5f 80 c7 c6 f4
0x0050 03 88 c3 0a d4 06 ab 05
0x0058 13 dc d6 f9 fd 73 76 56
0x0060 28 6e 11 77 d0 0f 88 8a
0x0068 db 31 c4And here's the same data, split up by octet and meaning so you can see what each piece does 🦋
0x0000 d2 packet type: SEIPD
0x0001 69 packet length
0x0002 02 v2 SEIPD
0x0003 07 cipher: AES128
0x0004 02 AEAD mode: OCB
0x0005 06 chunk size (2^12 octets)
0x0006 61 64 salt
0x0008 16 53 5b e0 b0 71 6d 60
0x0010 e0 52 a5 6c 4c 40 7f 9e
0x0018 b3 6b 0e fa fe 9a d0 a0
0x0020 df 9b 03 3c 69 a2
0x0026 1b a9 chunk #0 encrypted data
0x0028 eb d2 c0 ec 95 bf 56 9d
0x0030 25 c9 99 ee 4a 3d e1 70
0x0038 58 f4 0d fa 8b 4c 68 2b
0x0040 e3 fb bb d7 b2 7e b0 f5
0x0048 9b b5 00
0x004b 5f 80 c7 c6 f4 chunk #0 AEAD tag
0x0050 03 88 c3 0a d4 06 ab 05
0x0058 13 dc d6
0x005b f9 fd 73 76 56 final AEAD tag (#1)
S0x0060 28 6e 11 77 d0 0f 88 8a
0x0068 db 31 c4A.8.4. Decryption of Data 🔓
Time to decrypt it with the session key 💜 Starting AEAD-OCB decryption of the data now.
HKDF info:
d2 02 07 02 06HKDF output:
45 12 f7 14 9d 86 33 41 52 7c 65 67 d5 bf fc 42
5f af 32 50 21 2f f9Message key:
45 12 f7 14 9d 86 33 41 52 7c 65 67 d5 bf fc 42Initialization vector:
5f af 32 50 21 2f f9Chunk #0:
Nonce:
5f af 32 50 21 2f f9 00 00 00 00 00 00 00 00Additional authenticated data:
d2 02 07 02 06Encrypted data chunk:
1b a9 eb d2 c0 ec 95 bf 56 9d 25 c9 99 ee 4a 3d
e1 70 58 f4 0d fa 8b 4c 68 2b e3 fb bb d7 b2 7e
b0 f5 9b b5 00 5f 80 c7 c6 f4 03 88 c3 0a d4 06
ab 05 13 dc d6Chunk #0 is decrypted, slay 🙌
Literal Data packet with the string contents Hello, world!:
cb 13 62 00 00 00 00 00 48 65 6c 6c 6f 2c 20 77
6f 72 6c 64 21Padding packet:
d5 0e c5 a2 93 07 29 91 62 81 47 d7 2c 8f 86 b7Checking the final tag 🔍
Final nonce:
5f af 32 50 21 2f f9 00 00 00 00 00 00 00 01Final additional authenticated data:
d2 02 07 02 06 00 00 00 00 00 00 00 25A.8.5. Complete X25519-AEAD-OCB Encrypted Packet Sequence 🎁
Here's the whole thing wrapped up as one ASCII-armored message, ready to go 💌
-----BEGIN PGP MESSAGE-----
wV0GIQYSyD8ecG9jCP4VGkF3Q6HwM3kOk+mXhIjR2zeNqZMIhRmHzxjV8bU/gXzO
WgBM85PMiVi93AZfJfhK9QmxfdNnZBjeo1VDeVZheQHgaVf7yopqR6W1FT6NOrfS
aQIHAgZhZBZTW+CwcW1g4FKlbExAf56zaw76/prQoN+bAzxpohup69LA7JW/Vp0l
yZnuSj3hcFj0DfqLTGgr4/u717J+sPWbtQBfgMfG9AOIwwrUBqsFE9zW+f1zdlYo
bhF30A+IitsxxA==
-----END PGP MESSAGE-----A.9. Sample AEAD-EAX Encryption and Decryption 🌈
This one encrypts the plain string Hello, world! with the passphrase password, using AES-128 with AEAD-EAX encryption. Same vibe, different mode 💅
A.9.1. Sample Version 6 Symmetric Key Encrypted Session Key Packet 🧚
Here are the raw octets for this packet:
0x0000 c3 40 06 1e 07 01 0b 03
0x0008 08 a5 ae 57 9d 1f c5 d8
0x0010 2b ff 69 22 4f 91 99 93
0x0018 b3 50 6f a3 b5 9a 6a 73
0x0020 cf f8 c5 ef c5 f4 1c 57
0x0028 fb 54 e1 c2 26 81 5d 78
0x0030 28 f5 f9 2c 45 4e b6 5e
0x0038 be 00 ab 59 86 c6 8e 6e
0x0040 7c 55And the same data, broken down by octet and meaning:
0x0000 c3 packet type: SKESK
0x0001 40 packet length
0x0002 06 v6 SKESK
0x0003 1e length through end of AEAD nonce
0x0004 07 cipher: AES128
0x0005 01 AEAD mode: EAX
0x0006 0b length of S2K
0x0007 03 S2K type: iterated+salted
0x0008 08 S2K hash: SHA2-256
0x0009 a5 ae 57 9d 1f c5 d8 S2K salt
0x0010 2b
0x0011 ff S2K iterations (65011712 octets)
0x0012 69 22 4f 91 99 93 AEAD nonce
0x0018 b3 50 6f a3 b5 9a 6a 73
0x0020 cf f8
0x0022 c5 ef c5 f4 1c 57 encrypted session key
0x0028 fb 54 e1 c2 26 81 5d 78
0x0030 28 f5
0x0032 f9 2c 45 4e b6 5e AEAD tag
0x0038 be 00 ab 59 86 c6 8e 6e
0x0040 7c 55A.9.2. Starting AEAD-EAX Decryption of the Session Key 🗝️
The derived key is:
15 49 67 e5 90 aa 1f 92 3e 1c 0a c6 4c 88 f2 3dHKDF info:
c3 06 07 01HKDF output:
2f ce 33 1f 39 dd 95 5c c4 1e 95 d8 70 c7 21 39Authenticated Data:
c3 06 07 01Nonce:
69 22 4f 91 99 93 b3 50 6f a3 b5 9a 6a 73 cf f8Decrypted session key:
38 81 ba fe 98 54 12 45 9b 86 c3 6f 98 cb 9a 5eA.9.3. Sample v2 SEIPD Packet 🎀
Here are the octets for this packet:
0x0000 d2 69 02 07 01 06 9f f9
0x0008 0e 3b 32 19 64 f3 a4 29
0x0010 13 c8 dc c6 61 93 25 01
0x0018 52 27 ef b7 ea ea a4 9f
0x0020 04 c2 e6 74 17 5d 4a 3d
0x0028 22 6e d6 af cb 9c a9 ac
0x0030 12 2c 14 70 e1 1c 63 d4
0x0038 c0 ab 24 1c 6a 93 8a d4
0x0040 8b f9 9a 5a 99 b9 0b ba
0x0048 83 25 de 61 04 75 40 25
0x0050 8a b7 95 9a 95 ad 05 1d
0x0058 da 96 eb 15 43 1d fe f5
0x0060 f5 e2 25 5c a7 82 61 54
0x0068 6e 33 9aNow the same data, broken out by octet and meaning:
0x0000 d2 packet type: SEIPD
0x0001 69 packet length
0x0002 02 v2 SEIPD
0x0003 07 cipher: AES128
0x0004 01 AEAD mode: EAX
0x0005 06 chunk size (2^12 octets)
0x0005 9f f9 salt
0x0008 0e 3b 32 19 64 f3 a4 29
0x0010 13 c8 dc c6 61 93 25 01
0x0018 52 27 ef b7 ea ea a4 9f
0x0020 04 c2 e6 74 17 5d
0x0026 4a 3d chunk #0 encrypted data
0x0028 22 6e d6 af cb 9c a9 ac
0x0030 12 2c 14 70 e1 1c 63 d4
0x0038 c0 ab 24 1c 6a 93 8a d4
0x0040 8b f9 9a 5a 99 b9 0b ba
0x0048 83 25 de
0x004b 61 04 75 40 25 chunk #0 AEAD tag
0x0050 8a b7 95 9a 95 ad 05 1d
0x0058 da 96 eb
0x005b 15 43 1d fe f5 final AEAD tag (#1)
0x0060 f5 e2 25 5c a7 82 61 54
0x0068 6e 33 9aA.9.4. Decryption of Data 🔓
Decrypting the data with the session key, using AEAD-EAX 🌸
HKDF info:
d2 02 07 01 06HKDF output:
b5 04 22 ac 1c 26 be 9d dd 83 1d 5b bb 36 b6 4f
78 b8 33 f2 e9 4a 60 c0Message key:
b5 04 22 ac 1c 26 be 9d dd 83 1d 5b bb 36 b6 4fInitialization vector:
78 b8 33 f2 e9 4a 60 c0Chunk #0:
Nonce:
78 b8 33 f2 e9 4a 60 c0 00 00 00 00 00 00 00 00Additional authenticated data:
d2 02 07 01 06Chunk #0 decrypted, periodt 💖
Literal Data packet with the string contents Hello, world!:
cb 13 62 00 00 00 00 00 48 65 6c 6c 6f 2c 20 77
6f 72 6c 64 21Padding packet:
d5 0e ae 5b f0 cd 67 05 50 03 55 81 6c b0 c8 ffChecking the final tag 🔍
Final nonce:
78 b8 33 f2 e9 4a 60 c0 00 00 00 00 00 00 00 01Final additional authenticated data:
d2 02 07 01 06 00 00 00 00 00 00 00 25A.9.5. Complete AEAD-EAX Encrypted Packet Sequence 🍓
The full EAX message, armored and ready to share 🫶
-----BEGIN PGP MESSAGE-----
w0AGHgcBCwMIpa5XnR/F2Cv/aSJPkZmTs1Bvo7WaanPP+MXvxfQcV/tU4cImgV14
KPX5LEVOtl6+AKtZhsaObnxV0mkCBwEGn/kOOzIZZPOkKRPI3MZhkyUBUifvt+rq
pJ8EwuZ0F11KPSJu1q/LnKmsEiwUcOEcY9TAqyQcapOK1Iv5mlqZuQu6gyXeYQR1
QCWKt5Wala0FHdqW6xVDHf719eIlXKeCYVRuM5o=
-----END PGP MESSAGE-----A.10. Sample AEAD-OCB Encryption and Decryption 👑
Last one in this batch! This example encrypts the plain string Hello, world! with the passphrase password, using AES-128 with AEAD-OCB encryption 🦋
A.10.1. Sample Version 6 Symmetric Key Encrypted Session Key Packet 🧚
Here are the raw octets:
0x0000 c3 3f 06 1d 07 02 0b 03
0x0008 08 56 a2 98 d2 f5 e3 64
0x0010 53 ff cf cc 5c 11 66 4e
0x0018 db 9d b4 25 90 d7 dc 46
0x0020 b0 72 41 b6 12 c3 81 2c
0x0028 ff fb ea 00 f2 34 7b 25
0x0030 64 11 23 f8 87 ae 60 d4
0x0038 fd 61 4e 08 37 d8 19 d3
0x0040 6cAnd the breakdown by octet and meaning:
0x0000 c3 packet type: SKESK
0x0001 3f packet length
0x0002 06 v6 SKESK
0x0003 1d length through end of AEAD nonce
0x0004 07 cipher: AES128
0x0005 02 AEAD mode: OCB
0x0006 0b length of S2K
0x0007 03 S2K type: iterated+salted
0x0008 08 S2K hash: SHA2-256
0x0009 56 a2 98 d2 f5 e3 64 S2K salt
0x0010 53
0x0011 ff S2K iterations (65011712 octets)
0x0012 cf cc 5c 11 66 4e AEAD nonce
0x0018 db 9d b4 25 90 d7 dc 46
0x0020 b0
0x0021 72 41 b6 12 c3 81 2c encrypted session key
0x0028 ff fb ea 00 f2 34 7b 25
0x0030 64
0x0031 11 23 f8 87 ae 60 d4 AEAD tag
0x0038 fd 61 4e 08 37 d8 19 d3
0x0040 6cA.10.2. Starting AEAD-OCB Decryption of the Session Key 🗝️
The derived key is:
e8 0d e2 43 a3 62 d9 3b 9d c6 07 ed e9 6a 73 56HKDF info:
c3 06 07 02HKDF output:
38 a9 b3 45 b5 68 0b b6 1b b6 5d 73 ee c7 ec d9Authenticated Data:
c3 06 07 02Nonce:
cf cc 5c 11 66 4e db 9d b4 25 90 d7 dc 46 b0Decrypted session key:
28 e7 9a b8 23 97 d3 c6 3d e2 4a c2 17 d7 b7 91A.10.3. Sample v2 SEIPD Packet 🎀
The octets for this packet are:
0x0000 d2 69 02 07 02 06 20 a6
0x0008 61 f7 31 fc 9a 30 32 b5
0x0010 62 33 26 02 7e 3a 5d 8d
0x0018 b5 74 8e be ff 0b 0c 59
0x0020 10 d0 9e cd d6 41 ff 9f
0x0028 d3 85 62 75 80 35 bc 49
0x0030 75 4c e1 bf 3f ff a7 da
0x0038 d0 a3 b8 10 4f 51 33 cf
0x0040 42 a4 10 0a 83 ee f4 ca
0x0048 1b 48 01 a8 84 6b f4 2b
0x0050 cd a7 c8 ce 9d 65 e2 12
0x0058 f3 01 cb cd 98 fd ca de
0x0060 69 4a 87 7a d4 24 73 23
0x0068 f6 e8 57And the same thing, broken out by octet and meaning:
0x0000 d2 packet type: SEIPD
0x0001 69 packet length
0x0002 02 v2 SEIPD
0x0003 07 cipher: AES128
0x0004 02 AEAD mode: OCB
0x0005 06 chunk size (2^12 octets)
0x0006 20 a6 salt
0x0008 61 f7 31 fc 9a 30 32 b5
0x0010 62 33 26 02 7e 3a 5d 8d
0x0018 b5 74 8e be ff 0b 0c 59
0x0020 10 d0 9e cd d6 41
0x0026 ff 9f chunk #0 encrypted data
0x0028 d3 85 62 75 80 35 bc 49
0x0030 75 4c e1 bf 3f ff a7 da
0x0038 d0 a3 b8 10 4f 51 33 cf
0x0040 42 a4 10 0a 83 ee f4 ca
0x0048 1b 48 01
0x004b a8 84 6b f4 2b chunk #0 authentication tag
0x0050 cd a7 c8 ce 9d 65 e2 12
0x0058 f3 01 cb
0x005b cd 98 fd ca de final AEAD tag (#1)
0x0060 69 4a 87 7a d4 24 73 23
0x0068 f6 e8 57A.10.4. Decryption of Data 💖
Okay so this is where the AEAD-OCB decryption actually happens, bestie 🫶 We're starting from the session key and walking through every single step, no cap. 🔥
HKDF info:
d2 02 07 02 06HKDF output:
71 66 2a 11 ee 5b 4e 08 14 4e 6d e8 83 a0 09 99
eb de 12 bb 57 0d cfMessage key:
71 66 2a 11 ee 5b 4e 08 14 4e 6d e8 83 a0 09 99Initialization vector:
eb de 12 bb 57 0d cfChunk #0:
Nonce:
eb de 12 bb 57 0d cf 00 00 00 00 00 00 00 00Additional authenticated data:
d2 02 07 02 06Decrypted chunk #0. ✨
Literal Data packet with the string contents Hello, world!:
cb 13 62 00 00 00 00 00 48 65 6c 6c 6f 2c 20 77
6f 72 6c 64 21Padding packet: 🧚
d5 0e ae 6a a1 64 9b 56 aa 83 5b 26 13 90 2b d2Authenticating final tag: 🦋
Final nonce:
eb de 12 bb 57 0d cf 00 00 00 00 00 00 00 01Final additional authenticated data:
d2 02 07 02 06 00 00 00 00 00 00 00 25A.10.5. Complete AEAD-OCB Encrypted Packet Sequence 🎀
Here's the whole thing, all wrapped up in one armored message, the full AEAD-OCB sequence start to finish. Main character energy 👑
-----BEGIN PGP MESSAGE-----
wz8GHQcCCwMIVqKY0vXjZFP/z8xcEWZO2520JZDX3EawckG2EsOBLP/76gDyNHsl
ZBEj+IeuYNT9YU4IN9gZ02zSaQIHAgYgpmH3MfyaMDK1YjMmAn46XY21dI6+/wsM
WRDQns3WQf+f04VidYA1vEl1TOG/P/+n2tCjuBBPUTPPQqQQCoPu9MobSAGohGv0
K82nyM6dZeIS8wHLzZj9yt5pSod61CRzI/boVw==
-----END PGP MESSAGE-----A.11. Sample AEAD-GCM Encryption and Decryption 🌸
Okay so this example takes the cleartext string Hello, world! and encrypts it with the passphrase password, using AES-128 with AEAD-GCM encryption. Let's gooo 💅
A.11.1. Sample Version 6 Symmetric Key Encrypted Session Key Packet 🦋
This packet contains the following series of octets:
0x0000 c3 3c 06 1a 07 03 0b 03
0x0008 08 e9 d3 97 85 b2 07 00
0x0010 08 ff b4 2e 7c 48 3e f4
0x0018 88 44 57 cb 37 26 b9 b3
0x0020 db 9f f7 76 e5 f4 d9 a4
0x0028 09 52 e2 44 72 98 85 1a
0x0030 bf ff 75 26 df 2d d5 54
0x0038 41 75 79 a7 79 9fThe same data, broken out by octet and semantics, is:
0x0000 c3 packet type: SKESK
0x0001 3c packet length
0x0002 06 v6 SKESK
0x0003 1a length through end of AEAD nonce
0x0004 07 cipher: AES128
0x0005 03 AEAD mode: GCM
0x0006 0b length of S2K
0x0007 03 S2K type: iterated+salted
0x0008 08 S2K hash: SHA2-256
0x0009 e9 d3 97 85 b2 07 00 S2K salt
0x0010 08
0x0011 ff S2K iterations (65011712 octets)
0x0012 b4 2e 7c 48 3e f4 AEAD nonce
0x0018 88 44 57 cb 37 26
0x001e b9 b3 encrypted session key
0x0020 db 9f f7 76 e5 f4 d9 a4
0x0028 09 52 e2 44 72 98
0x002e 85 1a AEAD tag
0x0030 bf ff 75 26 df 2d d5 54
0x0038 41 75 79 a7 79 9fA.11.2. Starting AEAD-GCM Decryption of the Session Key 🧚
The derived key is:
25 02 81 71 5b ba 78 28 ef 71 ef 64 c4 78 47 53HKDF info:
c3 06 07 03HKDF output:
7a 6f 9a b7 f9 9f 7e f8 db ef 84 1c 65 08 00 f5Authenticated Data:
c3 06 07 03Nonce:
b4 2e 7c 48 3e f4 88 44 57 cb 37 26Decrypted session key:
19 36 fc 85 68 98 02 74 bb 90 0d 83 19 36 0c 77A.11.3. Sample v2 SEIPD Packet 💜
This packet contains the following series of octets, is:
0x0000 d2 69 02 07 03 06 fc b9
0x0008 44 90 bc b9 8b bd c9 d1
0x0010 06 c6 09 02 66 94 0f 72
0x0018 e8 9e dc 21 b5 59 6b 15
0x0020 76 b1 01 ed 0f 9f fc 6f
0x0028 c6 d6 5b bf d2 4d cd 07
0x0030 90 96 6e 6d 1e 85 a3 00
0x0038 53 78 4c b1 d8 b6 a0 69
0x0040 9e f1 21 55 a7 b2 ad 62
0x0048 58 53 1b 57 65 1f d7 77
0x0050 79 12 fa 95 e3 5d 9b 40
0x0058 21 6f 69 a4 c2 48 db 28
0x0060 ff 43 31 f1 63 29 07 39
0x0068 9e 6f f9The same data, broken out by octet and semantics, is:
0x0000 d2 packet type: SEIPD
0x0001 69 packet length
0x0002 02 v2 SEIPD
0x0003 07 cipher: AES128
0x0004 03 AEAD mode: GCM
0x0005 06 chunk size (2^12 octets)
0x0006 fc b9 salt
0x0008 44 90 bc b9 8b bd c9 d1
0x0010 06 c6 09 02 66 94 0f 72
0x0018 e8 9e dc 21 b5 59 6b 15
0x0020 76 b1 01 ed 0f 9f
0x0026 fc 6f chunk #0 encrypted data
0x0028 c6 d6 5b bf d2 4d cd 07
0x0030 90 96 6e 6d 1e 85 a3 00
0x0038 53 78 4c b1 d8 b6 a0 69
0x0040 9e f1 21 55 a7 b2 ad 62
0x0048 58 53 1b
0x004b 57 65 1f d7 77 chunk #0 authentication tag
0x0050 79 12 fa 95 e3 5d 9b 40
0x0058 21 6f 69
0x005b a4 c2 48 db 28 final AEAD tag (#1)
0x0060 ff 43 31 f1 63 29 07 39
0x0068 9e 6f f9A.11.4. Decryption of Data 💅
Starting AEAD-GCM decryption of data, using the session key. Here we go, bestie 🫶
HKDF info:
d2 02 07 03 06HKDF output:
ea 14 38 80 3c b8 a4 77 40 ce 9b 54 c3 38 77 8d
4d 2b dc 2bMessage key:
ea 14 38 80 3c b8 a4 77 40 ce 9b 54 c3 38 77 8dInitialization vector:
4d 2b dc 2bChunk #0:
Nonce:
4d 2b dc 2b 00 00 00 00 00 00 00 00Additional authenticated data:
d2 02 07 03 06Decrypted chunk #0. ✨
Literal Data packet with the string contents Hello, world!:
cb 13 62 00 00 00 00 00 48 65 6c 6c 6f 2c 20 77
6f 72 6c 64 21Padding packet: 🌈
d5 0e 1c e2 26 9a 9e dd ef 81 03 21 72 b7 ed 7cAuthenticating final tag: 🦋
Final nonce:
4d 2b dc 2b 00 00 00 00 00 00 00 01Final additional authenticated data:
d2 02 07 03 06 00 00 00 00 00 00 00 25A.11.5. Complete AEAD-GCM Encrypted Packet Sequence 🔥
And here's the full GCM version, all packed up in one armored message. Iconic 💖
-----BEGIN PGP MESSAGE-----
wzwGGgcDCwMI6dOXhbIHAAj/tC58SD70iERXyzcmubPbn/d25fTZpAlS4kRymIUa
v/91Jt8t1VRBdXmneZ/SaQIHAwb8uUSQvLmLvcnRBsYJAmaUD3LontwhtVlrFXax
Ae0Pn/xvxtZbv9JNzQeQlm5tHoWjAFN4TLHYtqBpnvEhVaeyrWJYUxtXZR/Xd3kS
+pXjXZtAIW9ppMJI2yj/QzHxYykHOZ5v+Q==
-----END PGP MESSAGE-----A.12. Sample Messages Encrypted Using Argon2 🍓
These messages are the literal data Hello, world! encrypted using v1 SEIPD, with Argon2 and the passphrase "password", using different session key sizes. ngl this is so cool 🧚 In each example, the choice of symmetric cipher is the same in both the v4 SKESK packet and v1 SEIPD packet. In all cases, the Argon2 parameters are t = 1, p = 4, and m = 21.
A.12.1. V4 SKESK Using Argon2 with AES-128 💜
-----BEGIN PGP MESSAGE-----
Comment: Encrypted using AES with 128-bit key
Comment: Session key: 01FE16BBACFD1E7B78EF3B865187374F
wycEBwScUvg8J/leUNU1RA7N/zE2AQQVnlL8rSLPP5VlQsunlO+ECxHSPgGYGKY+
YJz4u6F+DDlDBOr5NRQXt/KJIf4m4mOlKyC/uqLbpnLJZMnTq3o79GxBTdIdOzhH
XfA3pqV4mTzF
-----END PGP MESSAGE-----A.12.2. V4 SKESK Using Argon2 with AES-192 🦋
-----BEGIN PGP MESSAGE-----
Comment: Encrypted using AES with 192-bit key
Comment: Session key: 27006DAE68E509022CE45A14E569E91001C2955...
Comment: Session key: ...AF8DFE194
wy8ECAThTKxHFTRZGKli3KNH4UP4AQQVhzLJ2va3FG8/pmpIPd/H/mdoVS5VBLLw
F9I+AdJ1Sw56PRYiKZjCvHg+2bnq02s33AJJoyBexBI4QKATFRkyez2gldJldRys
LVg77Mwwfgl2n/d572WciAM=
-----END PGP MESSAGE-----A.12.3. V4 SKESK Using Argon2 with AES-256 ✨
-----BEGIN PGP MESSAGE-----
Comment: Encrypted using AES with 256-bit key
Comment: Session key: BBEDA55B9AAE63DAC45D4F49D89DACF4AF37FEF...
Comment: Session key: ...C13BAB2F1F8E18FB74580D8B0
wzcECQS4eJUgIG/3mcaILEJFpmJ8AQQVnZ9l7KtagdClm9UaQ/Z6M/5roklSGpGu
623YmaXezGj80j4B+Ku1sgTdJo87X1Wrup7l0wJypZls21Uwd67m9koF60eefH/K
95D1usliXOEm8ayQJQmZrjf6K6v9PWwqMQ==
-----END PGP MESSAGE-----Appendix B. Upgrade Guidance (Adapting Implementations from RFCs 4880 and 6637) 🌸
This subsection offers a concise, non-normative summary of the substantial additions to and departures from [RFC4880] and [RFC6637]. It is intended to help implementers who are augmenting an existing implementation from those specifications to comply with this specification. Cryptographic algorithms marked with "MTI" are mandatory to implement. Okay this is the glow-up checklist, bestie 💅✨
- Public Key Signing Algorithms:
- Ed25519 (Sections 5.5.5.9 and 5.2.3.4) -- MTI
- Ed448 (Sections 5.5.5.10 and 5.2.3.5)
- EdDSALegacy with Ed25519Legacy (Sections 5.5.5.5 and 5.2.3.3)
- ECDSA with Brainpool curves (Section 9.2)
- Public Key Encryption Algorithms:
- X25519 (Sections 5.5.5.7 and 5.1.6) -- MTI
- X448 (Sections 5.5.5.8 and 5.1.7)
- ECDH with Curve25519Legacy (Section 9.2)
- ECDH with Brainpool curves (Section 9.2)
- AEAD Encryption:
- V2 SEIPD (Section 5.13.2)
- AEAD modes:
- OCB mode (Section 5.13.4) -- MTI
- EAX mode (Section 5.13.3)
- GCM mode (Section 5.13.5)
- V6 PKESK (Section 5.1.2)
- V6 SKESK (Section 5.3.2)
- Features signature subpacket: add flag for v2 SEIPD (Section 5.2.3.32)
- Signature Subpacket: Preferred AEAD Ciphersuites (Section 5.2.3.15)
- Secret key encryption: AEAD "S2K usage octet" (Sections 3.7.2 and 5.5.3)
- Version 6 Keys and Signatures:
- Version 6 Public Keys (Section 5.5.2.3)
- Version 6 Fingerprint and Key ID (Section 5.5.4.3)
- Version 6 Secret Keys (Section 5.5.3)
- Version 6 Signatures (Section 5.2.3)
- Version 6 One-Pass Signatures (Section 5.4)
- Certificate (Transferable Public Key) Structure:
- Preferences subpackets in Direct Key signatures (Section 5.2.3.10)
- Self-verifying revocation certificate (Section 10.1.2)
- User ID is explicitly optional (Section 10.1.1)
- S2K: Argon2 (Section 3.7.1.4)
- Subpacket: Intended Recipient Fingerprint (Section 5.2.3.36)
- Digest Algorithms: SHA3-256 and SHA3-512 (Section 9.5)
- Packet: Padding (Section 5.14)
- Message Structure: Packet Criticality (Section 4.3)
- Deprecations:
- Public Key Algorithms:
- Avoid RSA weak keys (Section 12.4)
- Avoid DSA (Section 12.5)
- Avoid ElGamal (Sections 12.6 and 5.1.4)
- For Version 6 Keys: Avoid EdDSA25519Legacy and Curve25519Legacy (Section 9.2)
- Digest Algorithms:
- Avoid MD5, SHA1, and RIPEMD160 (Section 9.5)
- Symmetric Key Algorithms:
- Avoid IDEA, TripleDES, and CAST5 (Section 9.3)
- S2K Specifier:
- Avoid Simple S2K (Section 3.7.1.1)
- Secret Key Protections (a.k.a. S2K Usage):
- Avoid MalleableCFB (Section 3.7.2.1)
- Packet Types:
- Avoid Symmetrically Encrypted Data (Sections 5.7 and 13.7)
- Literal Data Packet Metadata:
- Avoid Filename and Date fields (Section 5.9)
- Avoid Special _CONSOLE "filename" (Section 5.9.1)
- Packet Versions:
- Avoid Version 3 Public Keys (Section 5.5.2.1)
- Avoid Version 3 Signatures (Section 5.2)
- Signature Types:
- Avoid Reserved Signature Type ID 0xFF (Sections 5.2.1.16 and 5.2.4.1)
- Signature Subpackets:
- For Version 6 Signatures: Avoid Issuer Key ID (Section 5.2.3.12)
- Avoid Revocation Key (Section 5.2.3.23)
- ASCII Armor:
- Ignore; do not emit CRC (Section 6.1)
- Do not emit "Version" Armor Header (Section 6.2.2.1)
- Cleartext Signature Framework:
- Ignore; avoid emitting unnecessary Hash: headers (Section 6.2.2.3)
- Reject Cleartext Signature Framework signatures with invalid Hash: headers (Section 6.2.2.3) or any other Armor Header (Section 7.1)
- Public Key Algorithms:
B.1. Terminology Changes ✨
okay bestie, some of the words from older versions of the OpenPGP spec got a glow-up in this document 💅 you should know the old names so you're not lost when you read older stuff 🎀
back in the day these terms were floating around:
- "Radix-64" was the name for OpenPGP's ASCII Armor base64 encoding (Section 6). 🧁
- "Old packet format" meant the Legacy packet format (Section 4.2.2) that was around before [RFC2440]. 🕰️
- "New packet format" meant the OpenPGP packet format (Section 4.2.1) that got introduced in [RFC2440]. ✨
- "Certificate" was used ambiguously to mean multiple things. ngl that was chaos 💀 In this document, it means "Transferable Public Key" exclusively. 🔑
- "Preferred Symmetric Algorithms" was the old name for the "Preferred Symmetric Ciphers for v1 SEIPD" subpacket (Section 5.2.3.14). 💜
- "Modification Detection Code" or "MDC" was originally described as a distinct packet (Packet Type ID 19), and its corresponding flag in the Features signature subpacket (Section 5.2.3.32) was known as "Modification Detection". It is now described as an intrinsic part of v1 SEIPD (Section 5.13.1), and that same flag is known as "Version 1 Symmetrically Encrypted and Integrity Protected Data packet". 🌈 the glow-up is real
- "Packet Tag" was used to refer to the Packet Type ID (Section 5), or sometimes to the encoded Packet Type ID (Section 4.2). 🦋
Appendix C. Errata Addressed by This Document 🐛
okay so these are the verified errata that have been folded into this document or are otherwise sorted out here, and they're giving main character energy 👑
- [Errata-2199] - S2K hash/cipher octet correction
- [Errata-2200] - No implicit use of IDEA correction
- [Errata-2206] - PKESK acronym expansion
- [Errata-2208] - Signature key owner clarification
- [Errata-2214] - Signature hashing clarification
- [Errata-2216] - Self-signature applies to user ID correction
- [Errata-2219] - Session key encryption storage clarification
- [Errata-2222] - Simple hash MUST/MAY clarification
- [Errata-2226] - Native line endings SHOULD clarification
- [Errata-2234] - Radix-64/base64 clarification
- [Errata-2235] - ASCII/UTF-8 collation sequence clarification
- [Errata-2236] - Packet Composition is a sequence clarification
- [Errata-2238] - Subkey packets come after all User ID packets clarification
- [Errata-2240] - Subkey removal clarification
- [Errata-2242] - mL/emLen variable correction
- [Errata-2243] - CFB mode initialization vector (IV) clarification
- [Errata-2270] - SHA-224 octet sequence correction
- [Errata-2271] - Radix-64 correction
- [Errata-3298] - Key Revocation signatures correction
- [Errata-5491] - C code fix for CRC24_POLY define
- [Errata-7545] - Armor Header colon hex fix
- [Errata-7889] - Signature/certification correction
Acknowledgements 🫶
big thank-you hugs to the OpenPGP Design Team, who worked on this document and got it ready for working group consumption 💖 that's Stephen Farrell, Daniel Kahn Gillmor, Daniel Huigens, Jeffrey Lau, Yutaka Niibe, Justus Winter, and Paul Wouters 🌸
also thanks to Werner Koch for the early work on rfc4880bis, and to Andrey Jivsov for the work on [RFC6637] ✨
and this document also builds on tons of earlier work from so many other authors: Derek Atkins, Charles Breed, Dave Del Torto, Marc Dyksterhouse, Gail Haspert, Gene Hoffman, Paul Hoffman, Ben Laurie, Raph Levien, Colin Plumb, Will Price, Daphne Shaw, William Stallings, Mark Weaver, and Philip R. Zimmermann. they're all iconic 🙌
Authors' Addresses 📮
- Paul Wouters (editor)
- Aiven
- Email: paul.wouters@aiven.io
- Daniel Huigens
- Proton AG
- Email: d.huigens@protonmail.com
- Justus Winter
- Sequoia PGP
- Email: justus@sequoia-pgp.org
- Yutaka Niibe
- FSIJ
- Email: gniibe@fsij.org
📦 where this page comes from
- ietf: rfc9580 2024-07
- 💖 girlified by claude-haiku-5-5 on : a language model's rewrite of the original, under its terms