serval-dna/nacl/nacl-20110221/crypto_scalarmult/curve25519/athlon/smult.c

#include "crypto_scalarmult.h"

#define mult crypto_scalarmult_curve25519_athlon_mult
#define square crypto_scalarmult_curve25519_athlon_square

void crypto_scalarmult_curve25519_athlon_recip(double out[10],const double z[10])
{
  double z2[10];
  double z9[10];
  double z11[10];
  double z2_5_0[10];
  double z2_10_0[10];
  double z2_20_0[10];
  double z2_50_0[10];
  double z2_100_0[10];
  double t0[10];
  double t1[10];
  int i;

  /* 2 */ square(z2,z);
  /* 4 */ square(t1,z2);
  /* 8 */ square(t0,t1);
  /* 9 */ mult(z9,t0,z);
  /* 11 */ mult(z11,z9,z2);
  /* 22 */ square(t0,z11);
  /* 2^5 - 2^0 = 31 */ mult(z2_5_0,t0,z9);

  /* 2^6 - 2^1 */ square(t0,z2_5_0);
  /* 2^7 - 2^2 */ square(t1,t0);
  /* 2^8 - 2^3 */ square(t0,t1);
  /* 2^9 - 2^4 */ square(t1,t0);
  /* 2^10 - 2^5 */ square(t0,t1);
  /* 2^10 - 2^0 */ mult(z2_10_0,t0,z2_5_0);

  /* 2^11 - 2^1 */ square(t0,z2_10_0);
  /* 2^12 - 2^2 */ square(t1,t0);
  /* 2^20 - 2^10 */ for (i = 2;i < 10;i += 2) { square(t0,t1); square(t1,t0); }
  /* 2^20 - 2^0 */ mult(z2_20_0,t1,z2_10_0);

  /* 2^21 - 2^1 */ square(t0,z2_20_0);
  /* 2^22 - 2^2 */ square(t1,t0);
  /* 2^40 - 2^20 */ for (i = 2;i < 20;i += 2) { square(t0,t1); square(t1,t0); }
  /* 2^40 - 2^0 */ mult(t0,t1,z2_20_0);

  /* 2^41 - 2^1 */ square(t1,t0);
  /* 2^42 - 2^2 */ square(t0,t1);
  /* 2^50 - 2^10 */ for (i = 2;i < 10;i += 2) { square(t1,t0); square(t0,t1); }
  /* 2^50 - 2^0 */ mult(z2_50_0,t0,z2_10_0);

  /* 2^51 - 2^1 */ square(t0,z2_50_0);
  /* 2^52 - 2^2 */ square(t1,t0);
  /* 2^100 - 2^50 */ for (i = 2;i < 50;i += 2) { square(t0,t1); square(t1,t0); }
  /* 2^100 - 2^0 */ mult(z2_100_0,t1,z2_50_0);

  /* 2^101 - 2^1 */ square(t1,z2_100_0);
  /* 2^102 - 2^2 */ square(t0,t1);
  /* 2^200 - 2^100 */ for (i = 2;i < 100;i += 2) { square(t1,t0); square(t0,t1); }
  /* 2^200 - 2^0 */ mult(t1,t0,z2_100_0);

  /* 2^201 - 2^1 */ square(t0,t1);
  /* 2^202 - 2^2 */ square(t1,t0);
  /* 2^250 - 2^50 */ for (i = 2;i < 50;i += 2) { square(t0,t1); square(t1,t0); }
  /* 2^250 - 2^0 */ mult(t0,t1,z2_50_0);

  /* 2^251 - 2^1 */ square(t1,t0);
  /* 2^252 - 2^2 */ square(t0,t1);
  /* 2^253 - 2^3 */ square(t1,t0);
  /* 2^254 - 2^4 */ square(t0,t1);
  /* 2^255 - 2^5 */ square(t1,t0);
  /* 2^255 - 21 */ mult(out,t1,z11);
}

int crypto_scalarmult(unsigned char *q,
  const unsigned char *n,
  const unsigned char *p)
{
  double work[30];
  unsigned char e[32];
  int i;
  for (i = 0;i < 32;++i) e[i] = n[i];
  e[0] &= 248;
  e[31] &= 127;
  e[31] |= 64;
  crypto_scalarmult_curve25519_athlon_init();
  crypto_scalarmult_curve25519_athlon_todouble(work,p);
  crypto_scalarmult_curve25519_athlon_mainloop(work,e);
  crypto_scalarmult_curve25519_athlon_recip(work + 10,work + 10);
  mult(work + 20,work,work + 10);
  crypto_scalarmult_curve25519_athlon_fromdouble(q,work + 20);
  return 0;
}