release/src/router/libsodium/src/libsodium/crypto_scalarmult/curve25519/ref/smult_curve25519_ref.c

   1 /*
   2 version 20081011
   3 Matthew Dempsky
   4 Public domain.
   5 Derived from public domain code by D. J. Bernstein.
   6 */
   7
   8 #include "api.h"
   9
  10 #ifndef HAVE_TI_MODE
  11
  12 static void add(unsigned int out[32],const unsigned int a[32],const unsigned int b[32])
  13 {
  14   unsigned int j;
  15   unsigned int u;
  16   u = 0;
  17   for (j = 0;j < 31;++j) { u += a[j] + b[j]; out[j] = u & 255; u >>= 8; }
  18   u += a[31] + b[31]; out[31] = u;
  19 }
  20
  21 static void sub(unsigned int out[32],const unsigned int a[32],const unsigned int b[32])
  22 {
  23   unsigned int j;
  24   unsigned int u;
  25   u = 218;
  26   for (j = 0;j < 31;++j) {
  27     u += a[j] + 65280 - b[j];
  28     out[j] = u & 255;
  29     u >>= 8;
  30   }
  31   u += a[31] - b[31];
  32   out[31] = u;
  33 }
  34
  35 static void squeeze(unsigned int a[32])
  36 {
  37   unsigned int j;
  38   unsigned int u;
  39   u = 0;
  40   for (j = 0;j < 31;++j) { u += a[j]; a[j] = u & 255; u >>= 8; }
  41   u += a[31]; a[31] = u & 127;
  42   u = 19 * (u >> 7);
  43   for (j = 0;j < 31;++j) { u += a[j]; a[j] = u & 255; u >>= 8; }
  44   u += a[31]; a[31] = u;
  45 }
  46
  47 static const unsigned int minusp[32] = {
  48  19, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 128
  49 } ;
  50
  51 static void freeze(unsigned int a[32])
  52 {
  53   unsigned int aorig[32];
  54   unsigned int j;
  55   unsigned int negative;
  56
  57   for (j = 0;j < 32;++j) aorig[j] = a[j];
  58   add(a,a,minusp);
  59   negative = -((a[31] >> 7) & 1);
  60   for (j = 0;j < 32;++j) a[j] ^= negative & (aorig[j] ^ a[j]);
  61 }
  62
  63 static void mult(unsigned int out[32],const unsigned int a[32],const unsigned int b[32])
  64 {
  65   unsigned int i;
  66   unsigned int j;
  67   unsigned int u;
  68
  69   for (i = 0;i < 32;++i) {
  70     u = 0;
  71     for (j = 0;j <= i;++j) u += a[j] * b[i - j];
  72     for (j = i + 1;j < 32;++j) u += 38 * a[j] * b[i + 32 - j];
  73     out[i] = u;
  74   }
  75   squeeze(out);
  76 }
  77
  78 static void mult121665(unsigned int out[32],const unsigned int a[32])
  79 {
  80   unsigned int j;
  81   unsigned int u;
  82
  83   u = 0;
  84   for (j = 0;j < 31;++j) { u += 121665 * a[j]; out[j] = u & 255; u >>= 8; }
  85   u += 121665 * a[31]; out[31] = u & 127;
  86   u = 19 * (u >> 7);
  87   for (j = 0;j < 31;++j) { u += out[j]; out[j] = u & 255; u >>= 8; }
  88   u += out[j]; out[j] = u;
  89 }
  90
  91 static void square(unsigned int out[32],const unsigned int a[32])
  92 {
  93   unsigned int i;
  94   unsigned int j;
  95   unsigned int u;
  96
  97   for (i = 0;i < 32;++i) {
  98     u = 0;
  99     for (j = 0;j < i - j;++j) u += a[j] * a[i - j];
 100     for (j = i + 1;j < i + 32 - j;++j) u += 38 * a[j] * a[i + 32 - j];
 101     u *= 2;
 102     if ((i & 1) == 0) {
 103       u += a[i / 2] * a[i / 2];
 104       u += 38 * a[i / 2 + 16] * a[i / 2 + 16];
 105     }
 106     out[i] = u;
 107   }
 108   squeeze(out);
 109 }
 110
 111 static void select(unsigned int p[64],unsigned int q[64],const unsigned int r[64],const unsigned int s[64],unsigned int b)
 112 {
 113   unsigned int j;
 114   unsigned int t;
 115   unsigned int bminus1;
 116
 117   bminus1 = b - 1;
 118   for (j = 0;j < 64;++j) {
 119     t = bminus1 & (r[j] ^ s[j]);
 120     p[j] = s[j] ^ t;
 121     q[j] = r[j] ^ t;
 122   }
 123 }
 124
 125 static void mainloop(unsigned int work[64],const unsigned char e[32])
 126 {
 127   unsigned int xzm1[64];
 128   unsigned int xzm[64];
 129   unsigned int xzmb[64];
 130   unsigned int xzm1b[64];
 131   unsigned int xznb[64];
 132   unsigned int xzn1b[64];
 133   unsigned int a0[64];
 134   unsigned int a1[64];
 135   unsigned int b0[64];
 136   unsigned int b1[64];
 137   unsigned int c1[64];
 138   unsigned int r[32];
 139   unsigned int s[32];
 140   unsigned int t[32];
 141   unsigned int u[32];
 142   unsigned int j;
 143   unsigned int b;
 144   int pos;
 145
 146   for (j = 0;j < 32;++j) xzm1[j] = work[j];
 147   xzm1[32] = 1;
 148   for (j = 33;j < 64;++j) xzm1[j] = 0;
 149
 150   xzm[0] = 1;
 151   for (j = 1;j < 64;++j) xzm[j] = 0;
 152
 153   for (pos = 254;pos >= 0;--pos) {
 154     b = e[pos / 8] >> (pos & 7);
 155     b &= 1;
 156     select(xzmb,xzm1b,xzm,xzm1,b);
 157     add(a0,xzmb,xzmb + 32);
 158     sub(a0 + 32,xzmb,xzmb + 32);
 159     add(a1,xzm1b,xzm1b + 32);
 160     sub(a1 + 32,xzm1b,xzm1b + 32);
 161     square(b0,a0);
 162     square(b0 + 32,a0 + 32);
 163     mult(b1,a1,a0 + 32);
 164     mult(b1 + 32,a1 + 32,a0);
 165     add(c1,b1,b1 + 32);
 166     sub(c1 + 32,b1,b1 + 32);
 167     square(r,c1 + 32);
 168     sub(s,b0,b0 + 32);
 169     mult121665(t,s);
 170     add(u,t,b0);
 171     mult(xznb,b0,b0 + 32);
 172     mult(xznb + 32,s,u);
 173     square(xzn1b,c1);
 174     mult(xzn1b + 32,r,work);
 175     select(xzm,xzm1,xznb,xzn1b,b);
 176   }
 177
 178   for (j = 0;j < 64;++j) work[j] = xzm[j];
 179 }
 180
 181 static void recip(unsigned int out[32],const unsigned int z[32])
 182 {
 183   unsigned int z2[32];
 184   unsigned int z9[32];
 185   unsigned int z11[32];
 186   unsigned int z2_5_0[32];
 187   unsigned int z2_10_0[32];
 188   unsigned int z2_20_0[32];
 189   unsigned int z2_50_0[32];
 190   unsigned int z2_100_0[32];
 191   unsigned int t0[32];
 192   unsigned int t1[32];
 193   int i;
 194
 195   /* 2 */ square(z2,z);
 196   /* 4 */ square(t1,z2);
 197   /* 8 */ square(t0,t1);
 198   /* 9 */ mult(z9,t0,z);
 199   /* 11 */ mult(z11,z9,z2);
 200   /* 22 */ square(t0,z11);
 201   /* 2^5 - 2^0 = 31 */ mult(z2_5_0,t0,z9);
 202
 203   /* 2^6 - 2^1 */ square(t0,z2_5_0);
 204   /* 2^7 - 2^2 */ square(t1,t0);
 205   /* 2^8 - 2^3 */ square(t0,t1);
 206   /* 2^9 - 2^4 */ square(t1,t0);
 207   /* 2^10 - 2^5 */ square(t0,t1);
 208   /* 2^10 - 2^0 */ mult(z2_10_0,t0,z2_5_0);
 209
 210   /* 2^11 - 2^1 */ square(t0,z2_10_0);
 211   /* 2^12 - 2^2 */ square(t1,t0);
 212   /* 2^20 - 2^10 */ for (i = 2;i < 10;i += 2) { square(t0,t1); square(t1,t0); }
 213   /* 2^20 - 2^0 */ mult(z2_20_0,t1,z2_10_0);
 214
 215   /* 2^21 - 2^1 */ square(t0,z2_20_0);
 216   /* 2^22 - 2^2 */ square(t1,t0);
 217   /* 2^40 - 2^20 */ for (i = 2;i < 20;i += 2) { square(t0,t1); square(t1,t0); }
 218   /* 2^40 - 2^0 */ mult(t0,t1,z2_20_0);
 219
 220   /* 2^41 - 2^1 */ square(t1,t0);
 221   /* 2^42 - 2^2 */ square(t0,t1);
 222   /* 2^50 - 2^10 */ for (i = 2;i < 10;i += 2) { square(t1,t0); square(t0,t1); }
 223   /* 2^50 - 2^0 */ mult(z2_50_0,t0,z2_10_0);
 224
 225   /* 2^51 - 2^1 */ square(t0,z2_50_0);
 226   /* 2^52 - 2^2 */ square(t1,t0);
 227   /* 2^100 - 2^50 */ for (i = 2;i < 50;i += 2) { square(t0,t1); square(t1,t0); }
 228   /* 2^100 - 2^0 */ mult(z2_100_0,t1,z2_50_0);
 229
 230   /* 2^101 - 2^1 */ square(t1,z2_100_0);
 231   /* 2^102 - 2^2 */ square(t0,t1);
 232   /* 2^200 - 2^100 */ for (i = 2;i < 100;i += 2) { square(t1,t0); square(t0,t1); }
 233   /* 2^200 - 2^0 */ mult(t1,t0,z2_100_0);
 234
 235   /* 2^201 - 2^1 */ square(t0,t1);
 236   /* 2^202 - 2^2 */ square(t1,t0);
 237   /* 2^250 - 2^50 */ for (i = 2;i < 50;i += 2) { square(t0,t1); square(t1,t0); }
 238   /* 2^250 - 2^0 */ mult(t0,t1,z2_50_0);
 239
 240   /* 2^251 - 2^1 */ square(t1,t0);
 241   /* 2^252 - 2^2 */ square(t0,t1);
 242   /* 2^253 - 2^3 */ square(t1,t0);
 243   /* 2^254 - 2^4 */ square(t0,t1);
 244   /* 2^255 - 2^5 */ square(t1,t0);
 245   /* 2^255 - 21 */ mult(out,t1,z11);
 246 }
 247
 248 int crypto_scalarmult(unsigned char *q,
 249   const unsigned char *n,
 250   const unsigned char *p)
 251 {
 252   unsigned int work[96];
 253   unsigned char e[32];
 254   unsigned int i;
 255   for (i = 0;i < 32;++i) e[i] = n[i];
 256   e[0] &= 248;
 257   e[31] &= 127;
 258   e[31] |= 64;
 259   for (i = 0;i < 32;++i) work[i] = p[i];
 260   mainloop(work,e);
 261   recip(work + 32,work + 32);
 262   mult(work + 64,work,work + 32);
 263   freeze(work + 64);
 264   for (i = 0;i < 32;++i) q[i] = work[64 + i];
 265   return 0;
 266 }
 267
 268 #endif