// does NOT ignore top bit
internal static void fe_frombytes2(out FieldElement h, byte[] data, int offset)
{
Int64 h0 = load_4(data, offset);
Int64 h1 = load_3(data, offset + 4) << 6;
Int64 h2 = load_3(data, offset + 7) << 5;
Int64 h3 = load_3(data, offset + 10) << 3;
Int64 h4 = load_3(data, offset + 13) << 2;
Int64 h5 = load_4(data, offset + 16);
Int64 h6 = load_3(data, offset + 20) << 7;
Int64 h7 = load_3(data, offset + 23) << 5;
Int64 h8 = load_3(data, offset + 26) << 4;
Int64 h9 = load_3(data, offset + 29) << 2;
Int64 carry0;
Int64 carry1;
Int64 carry2;
Int64 carry3;
Int64 carry4;
Int64 carry5;
Int64 carry6;
Int64 carry7;
Int64 carry8;
Int64 carry9;
carry9 = (h9 + (Int64)(1 << 24)) >> 25; h0 += carry9 * 19; h9 -= carry9 << 25;
carry1 = (h1 + (Int64)(1 << 24)) >> 25; h2 += carry1; h1 -= carry1 << 25;
carry3 = (h3 + (Int64)(1 << 24)) >> 25; h4 += carry3; h3 -= carry3 << 25;
carry5 = (h5 + (Int64)(1 << 24)) >> 25; h6 += carry5; h5 -= carry5 << 25;
carry7 = (h7 + (Int64)(1 << 24)) >> 25; h8 += carry7; h7 -= carry7 << 25;
carry0 = (h0 + (Int64)(1 << 25)) >> 26; h1 += carry0; h0 -= carry0 << 26;
carry2 = (h2 + (Int64)(1 << 25)) >> 26; h3 += carry2; h2 -= carry2 << 26;
carry4 = (h4 + (Int64)(1 << 25)) >> 26; h5 += carry4; h4 -= carry4 << 26;
carry6 = (h6 + (Int64)(1 << 25)) >> 26; h7 += carry6; h6 -= carry6 << 26;
carry8 = (h8 + (Int64)(1 << 25)) >> 26; h9 += carry8; h8 -= carry8 << 26;
h.x0 = (int)h0;
h.x1 = (int)h1;
h.x2 = (int)h2;
h.x3 = (int)h3;
h.x4 = (int)h4;
h.x5 = (int)h5;
h.x6 = (int)h6;
h.x7 = (int)h7;
h.x8 = (int)h8;
h.x9 = (int)h9;
}
/*
* Replace (f,g) with (g,f) if b == 1;
* replace (f,g) with (f,g) if b == 0.
*
* Preconditions: b in {0,1}.
*/
public static void fe_cswap(ref FieldElement f, ref FieldElement g, uint b)
{
Int32 f0 = f.x0;
Int32 f1 = f.x1;
Int32 f2 = f.x2;
Int32 f3 = f.x3;
Int32 f4 = f.x4;
Int32 f5 = f.x5;
Int32 f6 = f.x6;
Int32 f7 = f.x7;
Int32 f8 = f.x8;
Int32 f9 = f.x9;
Int32 g0 = g.x0;
Int32 g1 = g.x1;
Int32 g2 = g.x2;
Int32 g3 = g.x3;
Int32 g4 = g.x4;
Int32 g5 = g.x5;
Int32 g6 = g.x6;
Int32 g7 = g.x7;
Int32 g8 = g.x8;
Int32 g9 = g.x9;
Int32 x0 = f0 ^ g0;
Int32 x1 = f1 ^ g1;
Int32 x2 = f2 ^ g2;
Int32 x3 = f3 ^ g3;
Int32 x4 = f4 ^ g4;
Int32 x5 = f5 ^ g5;
Int32 x6 = f6 ^ g6;
Int32 x7 = f7 ^ g7;
Int32 x8 = f8 ^ g8;
Int32 x9 = f9 ^ g9;
int negb = unchecked ((int)-b);
x0 &= negb;
x1 &= negb;
x2 &= negb;
x3 &= negb;
x4 &= negb;
x5 &= negb;
x6 &= negb;
x7 &= negb;
x8 &= negb;
x9 &= negb;
f.x0 = f0 ^ x0;
f.x1 = f1 ^ x1;
f.x2 = f2 ^ x2;
f.x3 = f3 ^ x3;
f.x4 = f4 ^ x4;
f.x5 = f5 ^ x5;
f.x6 = f6 ^ x6;
f.x7 = f7 ^ x7;
f.x8 = f8 ^ x8;
f.x9 = f9 ^ x9;
g.x0 = g0 ^ x0;
g.x1 = g1 ^ x1;
g.x2 = g2 ^ x2;
g.x3 = g3 ^ x3;
g.x4 = g4 ^ x4;
g.x5 = g5 ^ x5;
g.x6 = g6 ^ x6;
g.x7 = g7 ^ x7;
g.x8 = g8 ^ x8;
g.x9 = g9 ^ x9;
}
internal static void scalarmult(
out FieldElement q,
byte[] n, int noffset,
ref FieldElement p)
{
byte[] e = new byte[32]; //ToDo: remove allocation
UInt32 i;
FieldElement x1;
FieldElement x2;
FieldElement z2;
FieldElement x3;
FieldElement z3;
FieldElement tmp0;
FieldElement tmp1;
int pos;
UInt32 swap;
UInt32 b;
for (i = 0; i < 32; ++i)
{
e[i] = n[noffset + i];
}
ScalarOperations.sc_clamp(e, 0);
x1 = p;
FieldOperations.fe_1(out x2);
FieldOperations.fe_0(out z2);
x3 = x1;
FieldOperations.fe_1(out z3);
swap = 0;
for (pos = 254; pos >= 0; --pos)
{
b = (uint)(e[pos / 8] >> (pos & 7));
b &= 1;
swap ^= b;
FieldOperations.fe_cswap(ref x2, ref x3, swap);
FieldOperations.fe_cswap(ref z2, ref z3, swap);
swap = b;
/* qhasm: fe X2 */
/* qhasm: fe Z2 */
/* qhasm: fe X3 */
/* qhasm: fe Z3 */
/* qhasm: fe X4 */
/* qhasm: fe Z4 */
/* qhasm: fe X5 */
/* qhasm: fe Z5 */
/* qhasm: fe A */
/* qhasm: fe B */
/* qhasm: fe C */
/* qhasm: fe D */
/* qhasm: fe E */
/* qhasm: fe AA */
/* qhasm: fe BB */
/* qhasm: fe DA */
/* qhasm: fe CB */
/* qhasm: fe t0 */
/* qhasm: fe t1 */
/* qhasm: fe t2 */
/* qhasm: fe t3 */
/* qhasm: fe t4 */
/* qhasm: enter ladder */
/* qhasm: D = X3-Z3 */
/* asm 1: fe_sub(>D=fe#5,<X3=fe#3,<Z3=fe#4); */
/* asm 2: fe_sub(>D=tmp0,<X3=x3,<Z3=z3); */
FieldOperations.fe_sub(out tmp0, ref x3, ref z3);
/* qhasm: B = X2-Z2 */
/* asm 1: fe_sub(>B=fe#6,<X2=fe#1,<Z2=fe#2); */
/* asm 2: fe_sub(>B=tmp1,<X2=x2,<Z2=z2); */
FieldOperations.fe_sub(out tmp1, ref x2, ref z2);
/* qhasm: A = X2+Z2 */
/* asm 1: fe_add(>A=fe#1,<X2=fe#1,<Z2=fe#2); */
/* asm 2: fe_add(>A=x2,<X2=x2,<Z2=z2); */
FieldOperations.fe_add(out x2, ref x2, ref z2);
/* qhasm: C = X3+Z3 */
/* asm 1: fe_add(>C=fe#2,<X3=fe#3,<Z3=fe#4); */
/* asm 2: fe_add(>C=z2,<X3=x3,<Z3=z3); */
FieldOperations.fe_add(out z2, ref x3, ref z3);
/* qhasm: DA = D*A */
/* asm 1: fe_mul(>DA=fe#4,<D=fe#5,<A=fe#1); */
/* asm 2: fe_mul(>DA=z3,<D=tmp0,<A=x2); */
FieldOperations.fe_mul(out z3, ref tmp0, ref x2);
/* qhasm: CB = C*B */
/* asm 1: fe_mul(>CB=fe#2,<C=fe#2,<B=fe#6); */
/* asm 2: fe_mul(>CB=z2,<C=z2,<B=tmp1); */
FieldOperations.fe_mul(out z2, ref z2, ref tmp1);
/* qhasm: BB = B^2 */
/* asm 1: fe_sq(>BB=fe#5,<B=fe#6); */
/* asm 2: fe_sq(>BB=tmp0,<B=tmp1); */
FieldOperations.fe_sq(out tmp0, ref tmp1);
/* qhasm: AA = A^2 */
/* asm 1: fe_sq(>AA=fe#6,<A=fe#1); */
/* asm 2: fe_sq(>AA=tmp1,<A=x2); */
FieldOperations.fe_sq(out tmp1, ref x2);
/* qhasm: t0 = DA+CB */
/* asm 1: fe_add(>t0=fe#3,<DA=fe#4,<CB=fe#2); */
/* asm 2: fe_add(>t0=x3,<DA=z3,<CB=z2); */
FieldOperations.fe_add(out x3, ref z3, ref z2);
/* qhasm: assign x3 to t0 */
/* qhasm: t1 = DA-CB */
/* asm 1: fe_sub(>t1=fe#2,<DA=fe#4,<CB=fe#2); */
/* asm 2: fe_sub(>t1=z2,<DA=z3,<CB=z2); */
FieldOperations.fe_sub(out z2, ref z3, ref z2);
/* qhasm: X4 = AA*BB */
/* asm 1: fe_mul(>X4=fe#1,<AA=fe#6,<BB=fe#5); */
/* asm 2: fe_mul(>X4=x2,<AA=tmp1,<BB=tmp0); */
FieldOperations.fe_mul(out x2, ref tmp1, ref tmp0);
/* qhasm: E = AA-BB */
/* asm 1: fe_sub(>E=fe#6,<AA=fe#6,<BB=fe#5); */
/* asm 2: fe_sub(>E=tmp1,<AA=tmp1,<BB=tmp0); */
FieldOperations.fe_sub(out tmp1, ref tmp1, ref tmp0);
/* qhasm: t2 = t1^2 */
/* asm 1: fe_sq(>t2=fe#2,<t1=fe#2); */
/* asm 2: fe_sq(>t2=z2,<t1=z2); */
FieldOperations.fe_sq(out z2, ref z2);
/* qhasm: t3 = a24*E */
/* asm 1: fe_mul121666(>t3=fe#4,<E=fe#6); */
/* asm 2: fe_mul121666(>t3=z3,<E=tmp1); */
FieldOperations.fe_mul121666(out z3, ref tmp1);
/* qhasm: X5 = t0^2 */
/* asm 1: fe_sq(>X5=fe#3,<t0=fe#3); */
/* asm 2: fe_sq(>X5=x3,<t0=x3); */
FieldOperations.fe_sq(out x3, ref x3);
/* qhasm: t4 = BB+t3 */
/* asm 1: fe_add(>t4=fe#5,<BB=fe#5,<t3=fe#4); */
/* asm 2: fe_add(>t4=tmp0,<BB=tmp0,<t3=z3); */
FieldOperations.fe_add(out tmp0, ref tmp0, ref z3);
/* qhasm: Z5 = X1*t2 */
/* asm 1: fe_mul(>Z5=fe#4,x1,<t2=fe#2); */
/* asm 2: fe_mul(>Z5=z3,x1,<t2=z2); */
FieldOperations.fe_mul(out z3, ref x1, ref z2);
/* qhasm: Z4 = E*t4 */
/* asm 1: fe_mul(>Z4=fe#2,<E=fe#6,<t4=fe#5); */
/* asm 2: fe_mul(>Z4=z2,<E=tmp1,<t4=tmp0); */
FieldOperations.fe_mul(out z2, ref tmp1, ref tmp0);
/* qhasm: return */
}
FieldOperations.fe_cswap(ref x2, ref x3, swap);
FieldOperations.fe_cswap(ref z2, ref z3, swap);
FieldOperations.fe_invert(out z2, ref z2);
FieldOperations.fe_mul(out x2, ref x2, ref z2);
q = x2;
CryptoBytes.Wipe(e);
}
internal static void fe_pow22523(out FieldElement result, ref FieldElement z)
{
FieldElement t0;
FieldElement t1;
FieldElement t2;
int i;
/* qhasm: fe z1 */
/* qhasm: fe z2 */
/* qhasm: fe z8 */
/* qhasm: fe z9 */
/* qhasm: fe z11 */
/* qhasm: fe z22 */
/* qhasm: fe z_5_0 */
/* qhasm: fe z_10_5 */
/* qhasm: fe z_10_0 */
/* qhasm: fe z_20_10 */
/* qhasm: fe z_20_0 */
/* qhasm: fe z_40_20 */
/* qhasm: fe z_40_0 */
/* qhasm: fe z_50_10 */
/* qhasm: fe z_50_0 */
/* qhasm: fe z_100_50 */
/* qhasm: fe z_100_0 */
/* qhasm: fe z_200_100 */
/* qhasm: fe z_200_0 */
/* qhasm: fe z_250_50 */
/* qhasm: fe z_250_0 */
/* qhasm: fe z_252_2 */
/* qhasm: fe z_252_3 */
/* qhasm: enter pow22523 */
/* qhasm: z2 = z1^2^1 */
/* asm 1: fe_sq(>z2=fe#1,<z1=fe#11); for (i = 1;i < 1;++i) fe_sq(>z2=fe#1,>z2=fe#1); */
/* asm 2: fe_sq(>z2=t0,<z1=z); for (i = 1;i < 1;++i) fe_sq(>z2=t0,>z2=t0); */
fe_sq(out t0, ref z); //for (i = 1; i < 1; ++i) fe_sq(out t0, ref t0);
/* qhasm: z8 = z2^2^2 */
/* asm 1: fe_sq(>z8=fe#2,<z2=fe#1); for (i = 1;i < 2;++i) fe_sq(>z8=fe#2,>z8=fe#2); */
/* asm 2: fe_sq(>z8=t1,<z2=t0); for (i = 1;i < 2;++i) fe_sq(>z8=t1,>z8=t1); */
fe_sq(out t1, ref t0); for (i = 1; i < 2; ++i)
{
fe_sq(out t1, ref t1);
}
/* qhasm: z9 = z1*z8 */
/* asm 1: fe_mul(>z9=fe#2,<z1=fe#11,<z8=fe#2); */
/* asm 2: fe_mul(>z9=t1,<z1=z,<z8=t1); */
fe_mul(out t1, ref z, ref t1);
/* qhasm: z11 = z2*z9 */
/* asm 1: fe_mul(>z11=fe#1,<z2=fe#1,<z9=fe#2); */
/* asm 2: fe_mul(>z11=t0,<z2=t0,<z9=t1); */
fe_mul(out t0, ref t0, ref t1);
/* qhasm: z22 = z11^2^1 */
/* asm 1: fe_sq(>z22=fe#1,<z11=fe#1); for (i = 1;i < 1;++i) fe_sq(>z22=fe#1,>z22=fe#1); */
/* asm 2: fe_sq(>z22=t0,<z11=t0); for (i = 1;i < 1;++i) fe_sq(>z22=t0,>z22=t0); */
fe_sq(out t0, ref t0); //for (i = 1; i < 1; ++i) fe_sq(out t0, ref t0);
/* qhasm: z_5_0 = z9*z22 */
/* asm 1: fe_mul(>z_5_0=fe#1,<z9=fe#2,<z22=fe#1); */
/* asm 2: fe_mul(>z_5_0=t0,<z9=t1,<z22=t0); */
fe_mul(out t0, ref t1, ref t0);
/* qhasm: z_10_5 = z_5_0^2^5 */
/* asm 1: fe_sq(>z_10_5=fe#2,<z_5_0=fe#1); for (i = 1;i < 5;++i) fe_sq(>z_10_5=fe#2,>z_10_5=fe#2); */
/* asm 2: fe_sq(>z_10_5=t1,<z_5_0=t0); for (i = 1;i < 5;++i) fe_sq(>z_10_5=t1,>z_10_5=t1); */
fe_sq(out t1, ref t0); for (i = 1; i < 5; ++i)
{
fe_sq(out t1, ref t1);
}
/* qhasm: z_10_0 = z_10_5*z_5_0 */
/* asm 1: fe_mul(>z_10_0=fe#1,<z_10_5=fe#2,<z_5_0=fe#1); */
/* asm 2: fe_mul(>z_10_0=t0,<z_10_5=t1,<z_5_0=t0); */
fe_mul(out t0, ref t1, ref t0);
/* qhasm: z_20_10 = z_10_0^2^10 */
/* asm 1: fe_sq(>z_20_10=fe#2,<z_10_0=fe#1); for (i = 1;i < 10;++i) fe_sq(>z_20_10=fe#2,>z_20_10=fe#2); */
/* asm 2: fe_sq(>z_20_10=t1,<z_10_0=t0); for (i = 1;i < 10;++i) fe_sq(>z_20_10=t1,>z_20_10=t1); */
fe_sq(out t1, ref t0); for (i = 1; i < 10; ++i)
{
fe_sq(out t1, ref t1);
}
/* qhasm: z_20_0 = z_20_10*z_10_0 */
/* asm 1: fe_mul(>z_20_0=fe#2,<z_20_10=fe#2,<z_10_0=fe#1); */
/* asm 2: fe_mul(>z_20_0=t1,<z_20_10=t1,<z_10_0=t0); */
fe_mul(out t1, ref t1, ref t0);
/* qhasm: z_40_20 = z_20_0^2^20 */
/* asm 1: fe_sq(>z_40_20=fe#3,<z_20_0=fe#2); for (i = 1;i < 20;++i) fe_sq(>z_40_20=fe#3,>z_40_20=fe#3); */
/* asm 2: fe_sq(>z_40_20=t2,<z_20_0=t1); for (i = 1;i < 20;++i) fe_sq(>z_40_20=t2,>z_40_20=t2); */
fe_sq(out t2, ref t1); for (i = 1; i < 20; ++i)
{
fe_sq(out t2, ref t2);
}
/* qhasm: z_40_0 = z_40_20*z_20_0 */
/* asm 1: fe_mul(>z_40_0=fe#2,<z_40_20=fe#3,<z_20_0=fe#2); */
/* asm 2: fe_mul(>z_40_0=t1,<z_40_20=t2,<z_20_0=t1); */
fe_mul(out t1, ref t2, ref t1);
/* qhasm: z_50_10 = z_40_0^2^10 */
/* asm 1: fe_sq(>z_50_10=fe#2,<z_40_0=fe#2); for (i = 1;i < 10;++i) fe_sq(>z_50_10=fe#2,>z_50_10=fe#2); */
/* asm 2: fe_sq(>z_50_10=t1,<z_40_0=t1); for (i = 1;i < 10;++i) fe_sq(>z_50_10=t1,>z_50_10=t1); */
fe_sq(out t1, ref t1); for (i = 1; i < 10; ++i)
{
fe_sq(out t1, ref t1);
}
/* qhasm: z_50_0 = z_50_10*z_10_0 */
/* asm 1: fe_mul(>z_50_0=fe#1,<z_50_10=fe#2,<z_10_0=fe#1); */
/* asm 2: fe_mul(>z_50_0=t0,<z_50_10=t1,<z_10_0=t0); */
fe_mul(out t0, ref t1, ref t0);
/* qhasm: z_100_50 = z_50_0^2^50 */
/* asm 1: fe_sq(>z_100_50=fe#2,<z_50_0=fe#1); for (i = 1;i < 50;++i) fe_sq(>z_100_50=fe#2,>z_100_50=fe#2); */
/* asm 2: fe_sq(>z_100_50=t1,<z_50_0=t0); for (i = 1;i < 50;++i) fe_sq(>z_100_50=t1,>z_100_50=t1); */
fe_sq(out t1, ref t0); for (i = 1; i < 50; ++i)
{
fe_sq(out t1, ref t1);
}
/* qhasm: z_100_0 = z_100_50*z_50_0 */
/* asm 1: fe_mul(>z_100_0=fe#2,<z_100_50=fe#2,<z_50_0=fe#1); */
/* asm 2: fe_mul(>z_100_0=t1,<z_100_50=t1,<z_50_0=t0); */
fe_mul(out t1, ref t1, ref t0);
/* qhasm: z_200_100 = z_100_0^2^100 */
/* asm 1: fe_sq(>z_200_100=fe#3,<z_100_0=fe#2); for (i = 1;i < 100;++i) fe_sq(>z_200_100=fe#3,>z_200_100=fe#3); */
/* asm 2: fe_sq(>z_200_100=t2,<z_100_0=t1); for (i = 1;i < 100;++i) fe_sq(>z_200_100=t2,>z_200_100=t2); */
fe_sq(out t2, ref t1); for (i = 1; i < 100; ++i)
{
fe_sq(out t2, ref t2);
}
/* qhasm: z_200_0 = z_200_100*z_100_0 */
/* asm 1: fe_mul(>z_200_0=fe#2,<z_200_100=fe#3,<z_100_0=fe#2); */
/* asm 2: fe_mul(>z_200_0=t1,<z_200_100=t2,<z_100_0=t1); */
fe_mul(out t1, ref t2, ref t1);
/* qhasm: z_250_50 = z_200_0^2^50 */
/* asm 1: fe_sq(>z_250_50=fe#2,<z_200_0=fe#2); for (i = 1;i < 50;++i) fe_sq(>z_250_50=fe#2,>z_250_50=fe#2); */
/* asm 2: fe_sq(>z_250_50=t1,<z_200_0=t1); for (i = 1;i < 50;++i) fe_sq(>z_250_50=t1,>z_250_50=t1); */
fe_sq(out t1, ref t1); for (i = 1; i < 50; ++i)
{
fe_sq(out t1, ref t1);
}
/* qhasm: z_250_0 = z_250_50*z_50_0 */
/* asm 1: fe_mul(>z_250_0=fe#1,<z_250_50=fe#2,<z_50_0=fe#1); */
/* asm 2: fe_mul(>z_250_0=t0,<z_250_50=t1,<z_50_0=t0); */
fe_mul(out t0, ref t1, ref t0);
/* qhasm: z_252_2 = z_250_0^2^2 */
/* asm 1: fe_sq(>z_252_2=fe#1,<z_250_0=fe#1); for (i = 1;i < 2;++i) fe_sq(>z_252_2=fe#1,>z_252_2=fe#1); */
/* asm 2: fe_sq(>z_252_2=t0,<z_250_0=t0); for (i = 1;i < 2;++i) fe_sq(>z_252_2=t0,>z_252_2=t0); */
fe_sq(out t0, ref t0); for (i = 1; i < 2; ++i)
{
fe_sq(out t0, ref t0);
}
/* qhasm: z_252_3 = z_252_2*z1 */
/* asm 1: fe_mul(>z_252_3=fe#12,<z_252_2=fe#1,<z1=fe#11); */
/* asm 2: fe_mul(>z_252_3=out,<z_252_2=t0,<z1=z); */
fe_mul(out result, ref t0, ref z);
/* qhasm: return */
}