diff options
author | Daniel Baumann <daniel.baumann@progress-linux.org> | 2024-04-28 13:16:40 +0000 |
---|---|---|
committer | Daniel Baumann <daniel.baumann@progress-linux.org> | 2024-04-28 13:16:40 +0000 |
commit | 47ab3d4a42e9ab51c465c4322d2ec233f6324e6b (patch) | |
tree | a61a0ffd83f4a3def4b36e5c8e99630c559aa723 /src/crypto/ed25519/internal/edwards25519/scalarmult.go | |
parent | Initial commit. (diff) | |
download | golang-1.18-47ab3d4a42e9ab51c465c4322d2ec233f6324e6b.tar.xz golang-1.18-47ab3d4a42e9ab51c465c4322d2ec233f6324e6b.zip |
Adding upstream version 1.18.10.upstream/1.18.10upstream
Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>
Diffstat (limited to 'src/crypto/ed25519/internal/edwards25519/scalarmult.go')
-rw-r--r-- | src/crypto/ed25519/internal/edwards25519/scalarmult.go | 214 |
1 files changed, 214 insertions, 0 deletions
diff --git a/src/crypto/ed25519/internal/edwards25519/scalarmult.go b/src/crypto/ed25519/internal/edwards25519/scalarmult.go new file mode 100644 index 0000000..f7ca3ce --- /dev/null +++ b/src/crypto/ed25519/internal/edwards25519/scalarmult.go @@ -0,0 +1,214 @@ +// Copyright (c) 2019 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. + +package edwards25519 + +import "sync" + +// basepointTable is a set of 32 affineLookupTables, where table i is generated +// from 256i * basepoint. It is precomputed the first time it's used. +func basepointTable() *[32]affineLookupTable { + basepointTablePrecomp.initOnce.Do(func() { + p := NewGeneratorPoint() + for i := 0; i < 32; i++ { + basepointTablePrecomp.table[i].FromP3(p) + for j := 0; j < 8; j++ { + p.Add(p, p) + } + } + }) + return &basepointTablePrecomp.table +} + +var basepointTablePrecomp struct { + table [32]affineLookupTable + initOnce sync.Once +} + +// ScalarBaseMult sets v = x * B, where B is the canonical generator, and +// returns v. +// +// The scalar multiplication is done in constant time. +func (v *Point) ScalarBaseMult(x *Scalar) *Point { + basepointTable := basepointTable() + + // Write x = sum(x_i * 16^i) so x*B = sum( B*x_i*16^i ) + // as described in the Ed25519 paper + // + // Group even and odd coefficients + // x*B = x_0*16^0*B + x_2*16^2*B + ... + x_62*16^62*B + // + x_1*16^1*B + x_3*16^3*B + ... + x_63*16^63*B + // x*B = x_0*16^0*B + x_2*16^2*B + ... + x_62*16^62*B + // + 16*( x_1*16^0*B + x_3*16^2*B + ... + x_63*16^62*B) + // + // We use a lookup table for each i to get x_i*16^(2*i)*B + // and do four doublings to multiply by 16. + digits := x.signedRadix16() + + multiple := &affineCached{} + tmp1 := &projP1xP1{} + tmp2 := &projP2{} + + // Accumulate the odd components first + v.Set(NewIdentityPoint()) + for i := 1; i < 64; i += 2 { + basepointTable[i/2].SelectInto(multiple, digits[i]) + tmp1.AddAffine(v, multiple) + v.fromP1xP1(tmp1) + } + + // Multiply by 16 + tmp2.FromP3(v) // tmp2 = v in P2 coords + tmp1.Double(tmp2) // tmp1 = 2*v in P1xP1 coords + tmp2.FromP1xP1(tmp1) // tmp2 = 2*v in P2 coords + tmp1.Double(tmp2) // tmp1 = 4*v in P1xP1 coords + tmp2.FromP1xP1(tmp1) // tmp2 = 4*v in P2 coords + tmp1.Double(tmp2) // tmp1 = 8*v in P1xP1 coords + tmp2.FromP1xP1(tmp1) // tmp2 = 8*v in P2 coords + tmp1.Double(tmp2) // tmp1 = 16*v in P1xP1 coords + v.fromP1xP1(tmp1) // now v = 16*(odd components) + + // Accumulate the even components + for i := 0; i < 64; i += 2 { + basepointTable[i/2].SelectInto(multiple, digits[i]) + tmp1.AddAffine(v, multiple) + v.fromP1xP1(tmp1) + } + + return v +} + +// ScalarMult sets v = x * q, and returns v. +// +// The scalar multiplication is done in constant time. +func (v *Point) ScalarMult(x *Scalar, q *Point) *Point { + checkInitialized(q) + + var table projLookupTable + table.FromP3(q) + + // Write x = sum(x_i * 16^i) + // so x*Q = sum( Q*x_i*16^i ) + // = Q*x_0 + 16*(Q*x_1 + 16*( ... + Q*x_63) ... ) + // <------compute inside out--------- + // + // We use the lookup table to get the x_i*Q values + // and do four doublings to compute 16*Q + digits := x.signedRadix16() + + // Unwrap first loop iteration to save computing 16*identity + multiple := &projCached{} + tmp1 := &projP1xP1{} + tmp2 := &projP2{} + table.SelectInto(multiple, digits[63]) + + v.Set(NewIdentityPoint()) + tmp1.Add(v, multiple) // tmp1 = x_63*Q in P1xP1 coords + for i := 62; i >= 0; i-- { + tmp2.FromP1xP1(tmp1) // tmp2 = (prev) in P2 coords + tmp1.Double(tmp2) // tmp1 = 2*(prev) in P1xP1 coords + tmp2.FromP1xP1(tmp1) // tmp2 = 2*(prev) in P2 coords + tmp1.Double(tmp2) // tmp1 = 4*(prev) in P1xP1 coords + tmp2.FromP1xP1(tmp1) // tmp2 = 4*(prev) in P2 coords + tmp1.Double(tmp2) // tmp1 = 8*(prev) in P1xP1 coords + tmp2.FromP1xP1(tmp1) // tmp2 = 8*(prev) in P2 coords + tmp1.Double(tmp2) // tmp1 = 16*(prev) in P1xP1 coords + v.fromP1xP1(tmp1) // v = 16*(prev) in P3 coords + table.SelectInto(multiple, digits[i]) + tmp1.Add(v, multiple) // tmp1 = x_i*Q + 16*(prev) in P1xP1 coords + } + v.fromP1xP1(tmp1) + return v +} + +// basepointNafTable is the nafLookupTable8 for the basepoint. +// It is precomputed the first time it's used. +func basepointNafTable() *nafLookupTable8 { + basepointNafTablePrecomp.initOnce.Do(func() { + basepointNafTablePrecomp.table.FromP3(NewGeneratorPoint()) + }) + return &basepointNafTablePrecomp.table +} + +var basepointNafTablePrecomp struct { + table nafLookupTable8 + initOnce sync.Once +} + +// VarTimeDoubleScalarBaseMult sets v = a * A + b * B, where B is the canonical +// generator, and returns v. +// +// Execution time depends on the inputs. +func (v *Point) VarTimeDoubleScalarBaseMult(a *Scalar, A *Point, b *Scalar) *Point { + checkInitialized(A) + + // Similarly to the single variable-base approach, we compute + // digits and use them with a lookup table. However, because + // we are allowed to do variable-time operations, we don't + // need constant-time lookups or constant-time digit + // computations. + // + // So we use a non-adjacent form of some width w instead of + // radix 16. This is like a binary representation (one digit + // for each binary place) but we allow the digits to grow in + // magnitude up to 2^{w-1} so that the nonzero digits are as + // sparse as possible. Intuitively, this "condenses" the + // "mass" of the scalar onto sparse coefficients (meaning + // fewer additions). + + basepointNafTable := basepointNafTable() + var aTable nafLookupTable5 + aTable.FromP3(A) + // Because the basepoint is fixed, we can use a wider NAF + // corresponding to a bigger table. + aNaf := a.nonAdjacentForm(5) + bNaf := b.nonAdjacentForm(8) + + // Find the first nonzero coefficient. + i := 255 + for j := i; j >= 0; j-- { + if aNaf[j] != 0 || bNaf[j] != 0 { + break + } + } + + multA := &projCached{} + multB := &affineCached{} + tmp1 := &projP1xP1{} + tmp2 := &projP2{} + tmp2.Zero() + + // Move from high to low bits, doubling the accumulator + // at each iteration and checking whether there is a nonzero + // coefficient to look up a multiple of. + for ; i >= 0; i-- { + tmp1.Double(tmp2) + + // Only update v if we have a nonzero coeff to add in. + if aNaf[i] > 0 { + v.fromP1xP1(tmp1) + aTable.SelectInto(multA, aNaf[i]) + tmp1.Add(v, multA) + } else if aNaf[i] < 0 { + v.fromP1xP1(tmp1) + aTable.SelectInto(multA, -aNaf[i]) + tmp1.Sub(v, multA) + } + + if bNaf[i] > 0 { + v.fromP1xP1(tmp1) + basepointNafTable.SelectInto(multB, bNaf[i]) + tmp1.AddAffine(v, multB) + } else if bNaf[i] < 0 { + v.fromP1xP1(tmp1) + basepointNafTable.SelectInto(multB, -bNaf[i]) + tmp1.SubAffine(v, multB) + } + + tmp2.FromP1xP1(tmp1) + } + + v.fromP2(tmp2) + return v +} |