eme

README

EME for Go

EME (ECB-Mix-ECB or, clearer, Encrypt-Mix-Encrypt) is a wide-block encryption mode developed by Halevi and Rogaway in 2003 [eme].

EME uses multiple invocations of a block cipher to construct a new cipher of bigger block size (in multiples of 16 bytes, up to 2048 bytes).

Quoting from the original [eme] paper:

We describe a block-cipher mode of operation, EME, that turns an n-bit block cipher into a tweakable enciphering scheme that acts on strings of mn bits, where m ∈ [1..n]. The mode is parallelizable, but as serial-efficient as the non-parallelizable mode CMC [6]. EME can be used to solve the disk-sector encryption problem. The algorithm entails two layers of ECB encryption and a “lightweight mixing” in between. We prove EME secure, in the reduction-based sense of modern cryptography.

Figure 2 from the [eme] paper shows an overview of the transformation:

This is an implementation of EME in Go, complete with test vectors from IEEE [p1619-2] and Halevi [eme-32-testvec].

It has no dependencies outside the standard library.

Is it patentend?

In 2007, the UC Davis has decided to abandon [patabandon] the patent application [patappl] for EME.

Related algorithms

EME-32 is EME with the cipher set to AES and the length set to 512. That is, EME-32 [eme-32-pdf] is a subset of EME.

EME2, also known as EME* [emestar], is an extended version of EME that has built-in handling for data that is not a multiple of 16 bytes long.
EME2 has been selected for standardization in IEEE P1619.2 [p1619.2].

References

[eme]

A Parallelizable Enciphering Mode
Shai Halevi, Phillip Rogaway, 28 Jul 2003
https://eprint.iacr.org/2003/147.pdf

Note: This is the original EME paper. EME is specified for an arbitrary number of block-cipher blocks. EME-32 is a concrete implementation of EME with a fixed length of 32 AES blocks.

[eme-32-email]

Re: EME-32-AES with editorial comments
Shai Halevi, 07 Jun 2005
http://grouper.ieee.org/groups/1619/email/msg00310.html

[eme-32-pdf]

Draft Standard for Tweakable Wide-block Encryption
Shai Halevi, 02 June 2005
http://grouper.ieee.org/groups/1619/email/pdf00020.pdf

Note: This is the latest version of the EME-32 draft that I could find. It includes test vectors and C source code.

[eme-32-testvec]

Re: Test vectors for LRW and EME
Shai Halevi, 16 Nov 2004
http://grouper.ieee.org/groups/1619/email/msg00218.html

[emestar]

EME*: extending EME to handle arbitrary-length messages with associated data
Shai Halevi, 27 May 2004
https://eprint.iacr.org/2004/125.pdf

[patabandon]

Re: [P1619-2] Non-awareness patent statement made by UC Davis
Mat Ball, 26 Nov 2007
http://grouper.ieee.org/groups/1619/email-2/msg00005.html

[patappl]

Block cipher mode of operation for constructing a wide-blocksize block cipher from a conventional block cipher
US patent application US20040131182
http://www.google.com/patents/US20040131182

[p1619-2]

IEEE P1619.2™/D9 Draft Standard for Wide-Block Encryption for Shared Storage Media
IEEE, Dec 2008
http://siswg.net/index2.php?option=com_docman&task=doc_view&gid=156&Itemid=41

Note: This is a draft version. The final version is not freely available and must be bought from IEEE.

Package Changelog

v1.1.1, 2020-04-13

• Update go vet call in test.bash to work on recent Go versions
• No code changes

v1.1, 2017-03-05

• Add eme.New() / *EMECipher convenience wrapper
• Improve panic message and parameter wording

v1.0, 2015-12-08

• Stable release

Documentation

Overview

EME (ECB-Mix-ECB or, clearer, Encrypt-Mix-Encrypt) is a wide-block encryption mode developed by Halevi and Rogaway.

It was presented in the 2003 paper "A Parallelizable Enciphering Mode" by Halevi and Rogaway.

EME uses multiple invocations of a block cipher to construct a new cipher of bigger block size (in multiples of 16 bytes, up to 2048 bytes).

Index

• Constants
• func Transform(bc cipher.Block, tweak []byte, inputData []byte, direction directionConst) []byte
• type EMECipher
  • func New(bc cipher.Block) *EMECipher
  • func (e *EMECipher) Decrypt(tweak []byte, inputData []byte) []byte
  • func (e *EMECipher) Encrypt(tweak []byte, inputData []byte) []byte

Constants

const (
	// Encrypt "inputData"
	DirectionEncrypt = directionConst(true)
	// Decrypt "inputData"
	DirectionDecrypt = directionConst(false)
)

Functions

func Transform

func Transform(bc cipher.Block, tweak []byte, inputData []byte, direction directionConst) []byte

Transform - EME-encrypt or EME-decrypt, according to "direction" (defined in the constants DirectionEncrypt and DirectionDecrypt). The data in "inputData" is en- or decrypted with the block ciper "bc" under "tweak" (also known as IV).

The tweak is used to randomize the encryption in the same way as an IV. A use of this encryption mode envisioned by the authors of the algorithm was to encrypt each sector of a disk, with the tweak being the sector number. If you encipher the same data with the same tweak you will get the same ciphertext.

The result is returned in a freshly allocated slice of the same size as inputData.

Limitations: * The block cipher must have block size 16 (usually AES). * The size of "tweak" must be 16 * "inputData" must be a multiple of 16 bytes long If any of these pre-conditions are not met, the function will panic.

Note that you probably don't want to call this function directly and instead use eme.New(), which provides conventient wrappers.

Types

type EMECipher

type EMECipher struct {
	// contains filtered or unexported fields
}

EMECipher provides EME-Encryption and -Decryption functions that are more convenient than calling Transform directly.

func New

func New(bc cipher.Block) *EMECipher

New returns a new EMECipher object. "bc" must have a block size of 16, or subsequent calls to Encrypt and Decrypt will panic.

func (*EMECipher) Decrypt

func (e *EMECipher) Decrypt(tweak []byte, inputData []byte) []byte

Decrypt is equivalent to calling Transform with direction=DirectionDecrypt.

func (*EMECipher) Encrypt

func (e *EMECipher) Encrypt(tweak []byte, inputData []byte) []byte

Encrypt is equivalent to calling Transform with direction=DirectionEncrypt.