/*
 * Adapted from nacl-20110221/crypto_core/salsa20/ref/core.c:
 *
 * - Renamed ROUNDS -> crypto_core_ROUNDS.
 * - Added #include <sys/types.h> and used uint32_t rather than unsigned int.
 * - Rewrote body of core with QUARTERROUND.
 * - Replaced bytes by words; let caller align and order bytes.
 * - Merged xi += ji; out[i] = xi; into out[i] = ji + xi.
 * - Replaced Salsa20 by ChaCha, following
 *   <http://cr.yp.to/chacha/chacha-20080128.pdf>.
 */

/*
version 20080912
D. J. Bernstein
Public domain.
*/

#include <sys/types.h>

typedef uint32_t uint32;

#include "crypto_core.h"

static uint32 rotate(uint32 u,int c)
{
  return (u << c) | (u >> (32 - c));
}

#define	QUARTERROUND(a,b,c,d) do {					      \
	(a) += (b); (d) ^= (a); (d) = rotate((d), 16);			      \
	(c) += (d); (b) ^= (c); (b) = rotate((b), 12);			      \
	(a) += (b); (d) ^= (a); (d) = rotate((d),  8);			      \
	(c) += (d); (b) ^= (c); (b) = rotate((b),  7);			      \
} while (0)

void crypto_core(
        uint32 *out,
  const uint32 *in,
  const uint32 *k,
  const uint32 *c
)
{
  uint32 x0, x1, x2, x3, x4, x5, x6, x7, x8, x9, x10, x11, x12, x13, x14, x15;
  uint32 j0, j1, j2, j3, j4, j5, j6, j7, j8, j9, j10, j11, j12, j13, j14, j15;
  int i;

  j0 = x0 = c[0];
  j1 = x1 = c[1];
  j2 = x2 = c[2];
  j3 = x3 = c[3];
  j4 = x4 = k[0];
  j5 = x5 = k[1];
  j6 = x6 = k[2];
  j7 = x7 = k[3];
  j8 = x8 = k[4];
  j9 = x9 = k[5];
  j10 = x10 = k[6];
  j11 = x11 = k[7];
  j12 = x12 = in[0];
  j13 = x13 = in[1];
  j14 = x14 = in[2];
  j15 = x15 = in[3];

  for (i = crypto_core_ROUNDS;i > 0;i -= 2) {
    QUARTERROUND( x0, x4, x8,x12);
    QUARTERROUND( x1, x5, x9,x13);
    QUARTERROUND( x2, x6,x10,x14);
    QUARTERROUND( x3, x7,x11,x15);
    QUARTERROUND( x0, x5,x10,x15);
    QUARTERROUND( x1, x6,x11,x12);
    QUARTERROUND( x2, x7, x8,x13);
    QUARTERROUND( x3, x4, x9,x14);
  }

  out[0] = x0 + j0;
  out[1] = x1 + j1;
  out[2] = x2 + j2;
  out[3] = x3 + j3;
  out[4] = x4 + j4;
  out[5] = x5 + j5;
  out[6] = x6 + j6;
  out[7] = x7 + j7;
  out[8] = x8 + j8;
  out[9] = x9 + j9;
  out[10] = x10 + j10;
  out[11] = x11 + j11;
  out[12] = x12 + j12;
  out[13] = x13 + j13;
  out[14] = x14 + j14;
  out[15] = x15 + j15;
}