Files
crypto/sha3/keccakf.go
T
Péter Szilágyi 80b25ed484 crypto/sha3: use the assembly optimized KeccakF on amd64
NIST published a Keccak Code Package following the SHA-3 workshop
organized in 2014, containing optimized versions of various Keccak
functions for various architectures. This CL converts the GNU asm
code of the Keccak permutation for the x86_64 architecture into Go
assembly.

The code here is almost an identical copy of KeccakF1600_StatePermute,
with the only modification of converting the input state into the
implementation's internal representation and vice versa before return.
This keeps the algorithm an in-place version and avoids requiring
extra external state inits and data XORs before and after the permute.

The speed difference is:

benchmark                          old ns/op     new ns/op     delta
BenchmarkPermutationFunction-8     476           411           -13.66%
BenchmarkSha3_512_MTU-8            9910          8681          -12.40%
BenchmarkSha3_384_MTU-8            7124          6249          -12.28%
BenchmarkSha3_256_MTU-8            5666          4986          -12.00%
BenchmarkSha3_224_MTU-8            5401          4750          -12.05%
BenchmarkShake128_MTU-8            4614          3980          -13.74%
BenchmarkShake256_MTU-8            4935          4295          -12.97%
BenchmarkShake256_16x-8            71850         63798         -11.21%
BenchmarkShake256_1MiB-8           3784244       3285733       -13.17%
BenchmarkSha3_512_1MiB-8           7098875       6163359       -13.18%

benchmark                          old MB/s     new MB/s     speedup
BenchmarkPermutationFunction-8     420.11       486.35       1.16x
BenchmarkSha3_512_MTU-8            136.22       155.51       1.14x
BenchmarkSha3_384_MTU-8            189.49       216.03       1.14x
BenchmarkSha3_256_MTU-8            238.23       270.71       1.14x
BenchmarkSha3_224_MTU-8            249.91       284.19       1.14x
BenchmarkShake128_MTU-8            292.58       339.15       1.16x
BenchmarkShake256_MTU-8            273.53       314.28       1.15x
BenchmarkShake256_16x-8            228.03       256.81       1.13x
BenchmarkShake256_1MiB-8           277.09       319.13       1.15x
BenchmarkSha3_512_1MiB-8           147.71       170.13       1.15x

For further details, please see:
 - http://csrc.nist.gov/groups/ST/hash/sha-3/Aug2014/documents/vanassche_keccak_code.pdf
 - https://github.com/gvanas/KeccakCodePackage

Change-Id: I5b0b9395bba7d8c9acfe2b9c79f6e9c2cf858c7c
Reviewed-on: https://go-review.googlesource.com/17962
Reviewed-by: Adam Langley <agl@golang.org>
2016-09-13 20:24:05 +00:00

413 lines
9.7 KiB
Go
Raw Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
// Copyright 2014 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
// +build !amd64 appengine gccgo
package sha3
// rc stores the round constants for use in the ι step.
var rc = [24]uint64{
0x0000000000000001,
0x0000000000008082,
0x800000000000808A,
0x8000000080008000,
0x000000000000808B,
0x0000000080000001,
0x8000000080008081,
0x8000000000008009,
0x000000000000008A,
0x0000000000000088,
0x0000000080008009,
0x000000008000000A,
0x000000008000808B,
0x800000000000008B,
0x8000000000008089,
0x8000000000008003,
0x8000000000008002,
0x8000000000000080,
0x000000000000800A,
0x800000008000000A,
0x8000000080008081,
0x8000000000008080,
0x0000000080000001,
0x8000000080008008,
}
// keccakF1600 applies the Keccak permutation to a 1600b-wide
// state represented as a slice of 25 uint64s.
func keccakF1600(a *[25]uint64) {
// Implementation translated from Keccak-inplace.c
// in the keccak reference code.
var t, bc0, bc1, bc2, bc3, bc4, d0, d1, d2, d3, d4 uint64
for i := 0; i < 24; i += 4 {
// Combines the 5 steps in each round into 2 steps.
// Unrolls 4 rounds per loop and spreads some steps across rounds.
// Round 1
bc0 = a[0] ^ a[5] ^ a[10] ^ a[15] ^ a[20]
bc1 = a[1] ^ a[6] ^ a[11] ^ a[16] ^ a[21]
bc2 = a[2] ^ a[7] ^ a[12] ^ a[17] ^ a[22]
bc3 = a[3] ^ a[8] ^ a[13] ^ a[18] ^ a[23]
bc4 = a[4] ^ a[9] ^ a[14] ^ a[19] ^ a[24]
d0 = bc4 ^ (bc1<<1 | bc1>>63)
d1 = bc0 ^ (bc2<<1 | bc2>>63)
d2 = bc1 ^ (bc3<<1 | bc3>>63)
d3 = bc2 ^ (bc4<<1 | bc4>>63)
d4 = bc3 ^ (bc0<<1 | bc0>>63)
bc0 = a[0] ^ d0
t = a[6] ^ d1
bc1 = t<<44 | t>>(64-44)
t = a[12] ^ d2
bc2 = t<<43 | t>>(64-43)
t = a[18] ^ d3
bc3 = t<<21 | t>>(64-21)
t = a[24] ^ d4
bc4 = t<<14 | t>>(64-14)
a[0] = bc0 ^ (bc2 &^ bc1) ^ rc[i]
a[6] = bc1 ^ (bc3 &^ bc2)
a[12] = bc2 ^ (bc4 &^ bc3)
a[18] = bc3 ^ (bc0 &^ bc4)
a[24] = bc4 ^ (bc1 &^ bc0)
t = a[10] ^ d0
bc2 = t<<3 | t>>(64-3)
t = a[16] ^ d1
bc3 = t<<45 | t>>(64-45)
t = a[22] ^ d2
bc4 = t<<61 | t>>(64-61)
t = a[3] ^ d3
bc0 = t<<28 | t>>(64-28)
t = a[9] ^ d4
bc1 = t<<20 | t>>(64-20)
a[10] = bc0 ^ (bc2 &^ bc1)
a[16] = bc1 ^ (bc3 &^ bc2)
a[22] = bc2 ^ (bc4 &^ bc3)
a[3] = bc3 ^ (bc0 &^ bc4)
a[9] = bc4 ^ (bc1 &^ bc0)
t = a[20] ^ d0
bc4 = t<<18 | t>>(64-18)
t = a[1] ^ d1
bc0 = t<<1 | t>>(64-1)
t = a[7] ^ d2
bc1 = t<<6 | t>>(64-6)
t = a[13] ^ d3
bc2 = t<<25 | t>>(64-25)
t = a[19] ^ d4
bc3 = t<<8 | t>>(64-8)
a[20] = bc0 ^ (bc2 &^ bc1)
a[1] = bc1 ^ (bc3 &^ bc2)
a[7] = bc2 ^ (bc4 &^ bc3)
a[13] = bc3 ^ (bc0 &^ bc4)
a[19] = bc4 ^ (bc1 &^ bc0)
t = a[5] ^ d0
bc1 = t<<36 | t>>(64-36)
t = a[11] ^ d1
bc2 = t<<10 | t>>(64-10)
t = a[17] ^ d2
bc3 = t<<15 | t>>(64-15)
t = a[23] ^ d3
bc4 = t<<56 | t>>(64-56)
t = a[4] ^ d4
bc0 = t<<27 | t>>(64-27)
a[5] = bc0 ^ (bc2 &^ bc1)
a[11] = bc1 ^ (bc3 &^ bc2)
a[17] = bc2 ^ (bc4 &^ bc3)
a[23] = bc3 ^ (bc0 &^ bc4)
a[4] = bc4 ^ (bc1 &^ bc0)
t = a[15] ^ d0
bc3 = t<<41 | t>>(64-41)
t = a[21] ^ d1
bc4 = t<<2 | t>>(64-2)
t = a[2] ^ d2
bc0 = t<<62 | t>>(64-62)
t = a[8] ^ d3
bc1 = t<<55 | t>>(64-55)
t = a[14] ^ d4
bc2 = t<<39 | t>>(64-39)
a[15] = bc0 ^ (bc2 &^ bc1)
a[21] = bc1 ^ (bc3 &^ bc2)
a[2] = bc2 ^ (bc4 &^ bc3)
a[8] = bc3 ^ (bc0 &^ bc4)
a[14] = bc4 ^ (bc1 &^ bc0)
// Round 2
bc0 = a[0] ^ a[5] ^ a[10] ^ a[15] ^ a[20]
bc1 = a[1] ^ a[6] ^ a[11] ^ a[16] ^ a[21]
bc2 = a[2] ^ a[7] ^ a[12] ^ a[17] ^ a[22]
bc3 = a[3] ^ a[8] ^ a[13] ^ a[18] ^ a[23]
bc4 = a[4] ^ a[9] ^ a[14] ^ a[19] ^ a[24]
d0 = bc4 ^ (bc1<<1 | bc1>>63)
d1 = bc0 ^ (bc2<<1 | bc2>>63)
d2 = bc1 ^ (bc3<<1 | bc3>>63)
d3 = bc2 ^ (bc4<<1 | bc4>>63)
d4 = bc3 ^ (bc0<<1 | bc0>>63)
bc0 = a[0] ^ d0
t = a[16] ^ d1
bc1 = t<<44 | t>>(64-44)
t = a[7] ^ d2
bc2 = t<<43 | t>>(64-43)
t = a[23] ^ d3
bc3 = t<<21 | t>>(64-21)
t = a[14] ^ d4
bc4 = t<<14 | t>>(64-14)
a[0] = bc0 ^ (bc2 &^ bc1) ^ rc[i+1]
a[16] = bc1 ^ (bc3 &^ bc2)
a[7] = bc2 ^ (bc4 &^ bc3)
a[23] = bc3 ^ (bc0 &^ bc4)
a[14] = bc4 ^ (bc1 &^ bc0)
t = a[20] ^ d0
bc2 = t<<3 | t>>(64-3)
t = a[11] ^ d1
bc3 = t<<45 | t>>(64-45)
t = a[2] ^ d2
bc4 = t<<61 | t>>(64-61)
t = a[18] ^ d3
bc0 = t<<28 | t>>(64-28)
t = a[9] ^ d4
bc1 = t<<20 | t>>(64-20)
a[20] = bc0 ^ (bc2 &^ bc1)
a[11] = bc1 ^ (bc3 &^ bc2)
a[2] = bc2 ^ (bc4 &^ bc3)
a[18] = bc3 ^ (bc0 &^ bc4)
a[9] = bc4 ^ (bc1 &^ bc0)
t = a[15] ^ d0
bc4 = t<<18 | t>>(64-18)
t = a[6] ^ d1
bc0 = t<<1 | t>>(64-1)
t = a[22] ^ d2
bc1 = t<<6 | t>>(64-6)
t = a[13] ^ d3
bc2 = t<<25 | t>>(64-25)
t = a[4] ^ d4
bc3 = t<<8 | t>>(64-8)
a[15] = bc0 ^ (bc2 &^ bc1)
a[6] = bc1 ^ (bc3 &^ bc2)
a[22] = bc2 ^ (bc4 &^ bc3)
a[13] = bc3 ^ (bc0 &^ bc4)
a[4] = bc4 ^ (bc1 &^ bc0)
t = a[10] ^ d0
bc1 = t<<36 | t>>(64-36)
t = a[1] ^ d1
bc2 = t<<10 | t>>(64-10)
t = a[17] ^ d2
bc3 = t<<15 | t>>(64-15)
t = a[8] ^ d3
bc4 = t<<56 | t>>(64-56)
t = a[24] ^ d4
bc0 = t<<27 | t>>(64-27)
a[10] = bc0 ^ (bc2 &^ bc1)
a[1] = bc1 ^ (bc3 &^ bc2)
a[17] = bc2 ^ (bc4 &^ bc3)
a[8] = bc3 ^ (bc0 &^ bc4)
a[24] = bc4 ^ (bc1 &^ bc0)
t = a[5] ^ d0
bc3 = t<<41 | t>>(64-41)
t = a[21] ^ d1
bc4 = t<<2 | t>>(64-2)
t = a[12] ^ d2
bc0 = t<<62 | t>>(64-62)
t = a[3] ^ d3
bc1 = t<<55 | t>>(64-55)
t = a[19] ^ d4
bc2 = t<<39 | t>>(64-39)
a[5] = bc0 ^ (bc2 &^ bc1)
a[21] = bc1 ^ (bc3 &^ bc2)
a[12] = bc2 ^ (bc4 &^ bc3)
a[3] = bc3 ^ (bc0 &^ bc4)
a[19] = bc4 ^ (bc1 &^ bc0)
// Round 3
bc0 = a[0] ^ a[5] ^ a[10] ^ a[15] ^ a[20]
bc1 = a[1] ^ a[6] ^ a[11] ^ a[16] ^ a[21]
bc2 = a[2] ^ a[7] ^ a[12] ^ a[17] ^ a[22]
bc3 = a[3] ^ a[8] ^ a[13] ^ a[18] ^ a[23]
bc4 = a[4] ^ a[9] ^ a[14] ^ a[19] ^ a[24]
d0 = bc4 ^ (bc1<<1 | bc1>>63)
d1 = bc0 ^ (bc2<<1 | bc2>>63)
d2 = bc1 ^ (bc3<<1 | bc3>>63)
d3 = bc2 ^ (bc4<<1 | bc4>>63)
d4 = bc3 ^ (bc0<<1 | bc0>>63)
bc0 = a[0] ^ d0
t = a[11] ^ d1
bc1 = t<<44 | t>>(64-44)
t = a[22] ^ d2
bc2 = t<<43 | t>>(64-43)
t = a[8] ^ d3
bc3 = t<<21 | t>>(64-21)
t = a[19] ^ d4
bc4 = t<<14 | t>>(64-14)
a[0] = bc0 ^ (bc2 &^ bc1) ^ rc[i+2]
a[11] = bc1 ^ (bc3 &^ bc2)
a[22] = bc2 ^ (bc4 &^ bc3)
a[8] = bc3 ^ (bc0 &^ bc4)
a[19] = bc4 ^ (bc1 &^ bc0)
t = a[15] ^ d0
bc2 = t<<3 | t>>(64-3)
t = a[1] ^ d1
bc3 = t<<45 | t>>(64-45)
t = a[12] ^ d2
bc4 = t<<61 | t>>(64-61)
t = a[23] ^ d3
bc0 = t<<28 | t>>(64-28)
t = a[9] ^ d4
bc1 = t<<20 | t>>(64-20)
a[15] = bc0 ^ (bc2 &^ bc1)
a[1] = bc1 ^ (bc3 &^ bc2)
a[12] = bc2 ^ (bc4 &^ bc3)
a[23] = bc3 ^ (bc0 &^ bc4)
a[9] = bc4 ^ (bc1 &^ bc0)
t = a[5] ^ d0
bc4 = t<<18 | t>>(64-18)
t = a[16] ^ d1
bc0 = t<<1 | t>>(64-1)
t = a[2] ^ d2
bc1 = t<<6 | t>>(64-6)
t = a[13] ^ d3
bc2 = t<<25 | t>>(64-25)
t = a[24] ^ d4
bc3 = t<<8 | t>>(64-8)
a[5] = bc0 ^ (bc2 &^ bc1)
a[16] = bc1 ^ (bc3 &^ bc2)
a[2] = bc2 ^ (bc4 &^ bc3)
a[13] = bc3 ^ (bc0 &^ bc4)
a[24] = bc4 ^ (bc1 &^ bc0)
t = a[20] ^ d0
bc1 = t<<36 | t>>(64-36)
t = a[6] ^ d1
bc2 = t<<10 | t>>(64-10)
t = a[17] ^ d2
bc3 = t<<15 | t>>(64-15)
t = a[3] ^ d3
bc4 = t<<56 | t>>(64-56)
t = a[14] ^ d4
bc0 = t<<27 | t>>(64-27)
a[20] = bc0 ^ (bc2 &^ bc1)
a[6] = bc1 ^ (bc3 &^ bc2)
a[17] = bc2 ^ (bc4 &^ bc3)
a[3] = bc3 ^ (bc0 &^ bc4)
a[14] = bc4 ^ (bc1 &^ bc0)
t = a[10] ^ d0
bc3 = t<<41 | t>>(64-41)
t = a[21] ^ d1
bc4 = t<<2 | t>>(64-2)
t = a[7] ^ d2
bc0 = t<<62 | t>>(64-62)
t = a[18] ^ d3
bc1 = t<<55 | t>>(64-55)
t = a[4] ^ d4
bc2 = t<<39 | t>>(64-39)
a[10] = bc0 ^ (bc2 &^ bc1)
a[21] = bc1 ^ (bc3 &^ bc2)
a[7] = bc2 ^ (bc4 &^ bc3)
a[18] = bc3 ^ (bc0 &^ bc4)
a[4] = bc4 ^ (bc1 &^ bc0)
// Round 4
bc0 = a[0] ^ a[5] ^ a[10] ^ a[15] ^ a[20]
bc1 = a[1] ^ a[6] ^ a[11] ^ a[16] ^ a[21]
bc2 = a[2] ^ a[7] ^ a[12] ^ a[17] ^ a[22]
bc3 = a[3] ^ a[8] ^ a[13] ^ a[18] ^ a[23]
bc4 = a[4] ^ a[9] ^ a[14] ^ a[19] ^ a[24]
d0 = bc4 ^ (bc1<<1 | bc1>>63)
d1 = bc0 ^ (bc2<<1 | bc2>>63)
d2 = bc1 ^ (bc3<<1 | bc3>>63)
d3 = bc2 ^ (bc4<<1 | bc4>>63)
d4 = bc3 ^ (bc0<<1 | bc0>>63)
bc0 = a[0] ^ d0
t = a[1] ^ d1
bc1 = t<<44 | t>>(64-44)
t = a[2] ^ d2
bc2 = t<<43 | t>>(64-43)
t = a[3] ^ d3
bc3 = t<<21 | t>>(64-21)
t = a[4] ^ d4
bc4 = t<<14 | t>>(64-14)
a[0] = bc0 ^ (bc2 &^ bc1) ^ rc[i+3]
a[1] = bc1 ^ (bc3 &^ bc2)
a[2] = bc2 ^ (bc4 &^ bc3)
a[3] = bc3 ^ (bc0 &^ bc4)
a[4] = bc4 ^ (bc1 &^ bc0)
t = a[5] ^ d0
bc2 = t<<3 | t>>(64-3)
t = a[6] ^ d1
bc3 = t<<45 | t>>(64-45)
t = a[7] ^ d2
bc4 = t<<61 | t>>(64-61)
t = a[8] ^ d3
bc0 = t<<28 | t>>(64-28)
t = a[9] ^ d4
bc1 = t<<20 | t>>(64-20)
a[5] = bc0 ^ (bc2 &^ bc1)
a[6] = bc1 ^ (bc3 &^ bc2)
a[7] = bc2 ^ (bc4 &^ bc3)
a[8] = bc3 ^ (bc0 &^ bc4)
a[9] = bc4 ^ (bc1 &^ bc0)
t = a[10] ^ d0
bc4 = t<<18 | t>>(64-18)
t = a[11] ^ d1
bc0 = t<<1 | t>>(64-1)
t = a[12] ^ d2
bc1 = t<<6 | t>>(64-6)
t = a[13] ^ d3
bc2 = t<<25 | t>>(64-25)
t = a[14] ^ d4
bc3 = t<<8 | t>>(64-8)
a[10] = bc0 ^ (bc2 &^ bc1)
a[11] = bc1 ^ (bc3 &^ bc2)
a[12] = bc2 ^ (bc4 &^ bc3)
a[13] = bc3 ^ (bc0 &^ bc4)
a[14] = bc4 ^ (bc1 &^ bc0)
t = a[15] ^ d0
bc1 = t<<36 | t>>(64-36)
t = a[16] ^ d1
bc2 = t<<10 | t>>(64-10)
t = a[17] ^ d2
bc3 = t<<15 | t>>(64-15)
t = a[18] ^ d3
bc4 = t<<56 | t>>(64-56)
t = a[19] ^ d4
bc0 = t<<27 | t>>(64-27)
a[15] = bc0 ^ (bc2 &^ bc1)
a[16] = bc1 ^ (bc3 &^ bc2)
a[17] = bc2 ^ (bc4 &^ bc3)
a[18] = bc3 ^ (bc0 &^ bc4)
a[19] = bc4 ^ (bc1 &^ bc0)
t = a[20] ^ d0
bc3 = t<<41 | t>>(64-41)
t = a[21] ^ d1
bc4 = t<<2 | t>>(64-2)
t = a[22] ^ d2
bc0 = t<<62 | t>>(64-62)
t = a[23] ^ d3
bc1 = t<<55 | t>>(64-55)
t = a[24] ^ d4
bc2 = t<<39 | t>>(64-39)
a[20] = bc0 ^ (bc2 &^ bc1)
a[21] = bc1 ^ (bc3 &^ bc2)
a[22] = bc2 ^ (bc4 &^ bc3)
a[23] = bc3 ^ (bc0 &^ bc4)
a[24] = bc4 ^ (bc1 &^ bc0)
}
}