Data.SBV.Examples.Crypto.AES

An implementation of AES (Advanced Encryption Standard), using SBV. For details on AES, see http://en.wikipedia.org/wiki/Advanced_Encryption_Standard.

We do a T-box implementation, which leads to good C code as we can take advantage of look-up tables. Note that we make virtually no attempt to optimize our Haskell code. The concern here is not with getting Haskell running fast at all. The idea is to program the T-Box implementation as naturally and clearly as possible in Haskell, and have SBV's code-generator generate fast C code automatically. Therefore, we merely use ordinary Haskell lists as our data-structures, and do not bother with any unboxing or strictness annotations. Thus, we achieve the separation of concerns: Correctness via clairty and simplicity and proofs on the Haskell side, performance by relying on SBV's code generator. If necessary, the generated code can be FFI'd back into Haskell to complete the loop.

All 3 valid key sizes (128, 192, and 256) as required by the FIPS-197 standard are supported.

Formalizing GF(2^8)

type GF28 = SWord8

An element of the Galois Field 2^8, which are essentially polynomials with maximum degree 7. They are conveniently represented as values between 0 and 255.

gf28Mult :: GF28 -> GF28 -> GF28

Multiplication in GF(2^8). This is simple polynomial multipliation, followed by the irreducible polynomial x^8+x^4+x^3+x^1+1. We simply use the pMult function exported by SBV to do the operation.

gf28Pow :: GF28 -> Int -> GF28

Exponentiation by a constant in GF(2^8). The implementation uses the usual square-and-multiply trick to speed up the computation.

gf28Inverse :: GF28 -> GF28

Computing inverses in GF(2^8). By the mathematical properties of GF(2^8) and the particular irreducible polynomial used x^8+x^5+x^3+x^1+1, it turns out that raising to the 254 power gives us the multiplicative inverse. Of course, we can prove this using SBV:

>>> prove $ \x -> x ./= 0 ==> x `gf28Mult` gf28Inverse x .== 1
Q.E.D.

Note that we exclude 0 in our theorem, as it does not have a multiplicative inverse.

Implementing AES

Types and basic operations

type State = [SWord32]

AES state. The state consists of four 32-bit words, each of which is in turn treated as four GF28's, i.e., 4 bytes. The T-Box implementation keeps the four-bytes together for efficient representation.

type Key = [SWord32]

The key, which can be 128, 192, or 256 bits. Represented as a sequence of 32-bit words.

type KS = (Key, [Key], Key)

The key schedule. AES executes in rounds, and it treats first and last round keys slightly differently than the middle ones. We reflect that choice by being explicit about it in our type. The length of the middle list of keys depends on the key-size, which in turn determines the number of rounds.

toBytes :: SWord32 -> [GF28]

Conversion from 32-bit words to 4 constituent bytes.

fromBytes :: [GF28] -> SWord32

Conversion from 4 bytes, back to a 32-bit row, inverse of toBytes above. We have the following simple theorems stating this relationship formally:

>>> prove $ \a b c d -> toBytes (fromBytes [a, b, c, d]) .== [a, b, c, d]
Q.E.D.

>>> prove $ \r -> fromBytes (toBytes r) .== r
Q.E.D.

rotR :: [GF28] -> Int -> [GF28]

Rotating a state row by a fixed amount to the right.

The key schedule

roundConstants :: [GF28]

Definition of round-constants, as specified in Section 5.2 of the AES standard.

invMixColumns :: State -> State

The InvMixColumns transformation, as described in Section 5.3.3 of the standard. Note that this transformation is only used explicitly during key-expansion in the T-Box implementation of AES.

keyExpansion :: Int -> Key -> [Key]

Key expansion. Starting with the given key, returns an infinite sequence of words, as described by the AES standard, Section 5.2, Figure 11.

The S-box transformation

sboxTable :: [GF28]

The values of the AES S-box table. Note that we describe the S-box programmatically using the mathematical construction given in Section 5.1.1 of the standard. However, the code-generation will turn this into a mere look-up table, as it is just a constant table, all computation being done at "compile-time".

sbox :: GF28 -> GF28

The sbox transformation. We simply select from the sbox table. Note that we are obliged to give a default value (here 0) to be used if the index is out-of-bounds as required by SBV's select function. However, that will never happen since the table has all 256 elements in it.

The inverse S-box transformation

unSBoxTable :: [GF28]

The values of the inverse S-box table. Again, the construction is programmatic.

unSBox :: GF28 -> GF28

The inverse s-box transformation.

sboxInverseCorrect :: GF28 -> SBool

Prove that the sbox and unSBox are inverses. We have:

>>> prove sboxInverseCorrect
Q.E.D.

AddRoundKey transformation

addRoundKey :: Key -> State -> State

Adding the round-key to the current state. We simply exploit the fact that addition is just xor in implementing this transformation.

Tables for T-Box encryption

t0Func :: GF28 -> [GF28]

T-box table generation function for encryption

t0 :: GF28 -> SWord32

First look-up table used in encryption

t1 :: GF28 -> SWord32

Second look-up table used in encryption

t2 :: GF28 -> SWord32

Third look-up table used in encryption

t3 :: GF28 -> SWord32

Fourth look-up table used in encryption

Tables for T-Box decryption

u0Func :: GF28 -> [GF28]

T-box table generating function for decryption

u0 :: GF28 -> SWord32

First look-up table used in decryption

u1 :: GF28 -> SWord32

Second look-up table used in decryption

u2 :: GF28 -> SWord32

Third look-up table used in decryption

u3 :: GF28 -> SWord32

Fourth look-up table used in decryption

AES rounds

doRounds :: (Bool -> State -> Key -> State) -> KS -> State -> State

Generic round function. Given the function to perform one round, a key-schedule, and a starting state, it performs the AES rounds.

aesRound :: Bool -> State -> Key -> State

One encryption round. The first argument indicates whether this is the final round or not, in which case the construction is slightly different.

aesInvRound :: Bool -> State -> Key -> State

One decryption round. Similar to the encryption round, the first argument indicates whether this is the final round or not.

AES API

aesKeySchedule :: Key -> (KS, KS)

Key schedule. Given a 128, 192, or 256 bit key, expand it to get key-schedules for encryption and decryption. The key is given as a sequence of 32-bit words. (4 elements for 128-bits, 6 for 192, and 8 for 256.)

aesEncrypt :: [SWord32] -> KS -> [SWord32]

Block encryption. The first argument is the plain-text, which must have precisely 4 elements, for a total of 128-bits of input. The second argument is the key-schedule to be used, obtained by a call to aesKeySchedule. The output will always have 4 32-bit words, which is the cipher-text.

aesDecrypt :: [SWord32] -> KS -> [SWord32]

Block decryption. The arguments are the same as in aesEncrypt, except the first argument is the cipher-text and the output is the corresponding plain-text.

Test vectors

128-bit enc/dec test

t128Enc :: [SWord32]

128-bit encryption test, from Appendix C.1 of the AES standard:

>>> map hex8 t128Enc
["69c4e0d8","6a7b0430","d8cdb780","70b4c55a"]

t128Dec :: [SWord32]

128-bit decryption test, from Appendix C.1 of the AES standard:

>>> map hex8 t128Dec
["00112233","44556677","8899aabb","ccddeeff"]

192-bit enc/dec test

t192Enc :: [SWord32]

192-bit encryption test, from Appendix C.2 of the AES standard:

>>> map hex8 t192Enc
["dda97ca4","864cdfe0","6eaf70a0","ec0d7191"]

t192Dec :: [SWord32]

192-bit decryption test, from Appendix C.2 of the AES standard:

>>> map hex8 t192Dec
["00112233","44556677","8899aabb","ccddeeff"]

256-bit enc/dec test

t256Enc :: [SWord32]

256-bit encryption, from Appendix C.3 of the AES standard:

>>> map hex8 t256Enc
["8ea2b7ca","516745bf","eafc4990","4b496089"]

t256Dec :: [SWord32]

256-bit decryption, from Appendix C.3 of the AES standard:

>>> map hex8 t256Dec
["00112233","44556677","8899aabb","ccddeeff"]

Verification

While SMT based technologies can prove correct many small properties fairly quickly, it would be naive for them to automatically verify that our AES implementation is correct. (By correct, we mean decryption follewed by encryption yielding the same result.) However, we can state this property precisely using SBV, and use quick-check to gain some confidence.

aes128IsCorrect

Arguments

:: (SWord32, SWord32, SWord32, SWord32)

plain-text words

-> (SWord32, SWord32, SWord32, SWord32)

key-words

-> SBool

True if round-trip gives us plain-text back

Correctness theorem for 128-bit AES. Ideally, we would run:

  prove aes128IsCorrect

to get a proof automatically. Unfortunately, while SBV will successfully generate the proof obligation for this theorem and ship it to the SMT solver, it would be naive to expect the SMT-solver to finish that proof in any reasonable time with the currently available SMT solving technologies. Instead, we can issue:

  quickCheck aes128IsCorrect

and get some degree of confidence in our code. Similar predicates can be easily constructed for 192, and 256 bit cases as well.

Code generation

We have emphasized that our T-Box implementation in Haskell was guided by clarity and correctness, not performance. Indeed, our implementation is hardly the fastest AES implementation in Haskell. However, we can use it to automatically generate straight-line C-code that can run fairly fast.

For the purposes of illustration, we only show here how to generate code for a 128-bit AES block-encrypt function, that takes 8 32-bit words as an argument. The first 4 are the 128-bit input, and the final four are the 128-bit key. The impact of this is that the generated function would expand the key for each block of encryption, a needless task unless we change the key in every block. In a more serios application, we would instead generate code for both the aesKeySchedule and the aesEncrypt functions, thus reusing the key-schedule over many applications of the encryption call. (Unfortunately doing this is rather cumbersome right now, since Haskell does not support fixed-size lists.)

cgAES128BlockEncrypt :: IO ()

Code generation for 128-bit AES encryption.

The following sample from the generated code-lines show how T-Boxes are rendered as C arrays:

  static const SWord32 table1[] = {
      0xc66363a5UL, 0xf87c7c84UL, 0xee777799UL, 0xf67b7b8dUL,
      0xfff2f20dUL, 0xd66b6bbdUL, 0xde6f6fb1UL, 0x91c5c554UL,
      0x60303050UL, 0x02010103UL, 0xce6767a9UL, 0x562b2b7dUL,
      0xe7fefe19UL, 0xb5d7d762UL, 0x4dababe6UL, 0xec76769aUL,
      ...
      }

The generated program has 5 tables (one sbox table, and 4-Tboxes), all converted to fast C arrays. Here is a sample of the generated straightline C-code:

  const SWord8  s1915 = (SWord8) s1912;
  const SWord8  s1916 = table0[s1915];
  const SWord16 s1917 = (((SWord16) s1914) << 8) | ((SWord16) s1916);
  const SWord32 s1918 = (((SWord32) s1911) << 16) | ((SWord32) s1917);
  const SWord32 s1919 = s1844 ^ s1918;
  const SWord32 s1920 = s1903 ^ s1919;

The GNU C-compiler does a fine job of optimizing this straightline code to generate a fairly efficient C implementation.

C-library generation

The cgAES128BlockEncrypt example shows how to generate code for 128-bit AES encryption. As the generated function performs encryption on a given block, it performs key expansion as necessary. However, this is not quite practical: We would like to expand the key only once, and encrypt the stream of plain-text blocks using the same expanded key (potentially using some crypto-mode), until we decide to change the key. In this section, we show how to use SBV to instead generate a library of functions that can be used in such a scenario. The generated library is a typical .a archive, that can be linked using the C-compiler as usual.

aes128LibComponents :: [(String, SBVCodeGen ())]

Components of the AES-128 implementation that the library is generated from

cgAES128Library :: IO ()

Generate a C library, containing functions for performing 128-bit encdeckey-expansion. A note on performance: In a very rough speed test, the generated code was able to do 6.3 million block encryptions per second on a decent MacBook Pro. On the same machine, OpenSSL reports 8.2 million block encryptions per second. So, the generated code is about 25% slower as compared to the highly optimized OpenSSL implementation. (Note that the speed test was done somewhat simplistically, so these numbers should be considered very rough estimates.)

hex8 :: (SymWord a, Show a, Integral a) => SBV a -> String

For doctest purposes only