sgminer/kernel/fresh.cl

/*
 * FRESH kernel implementation.
 *
 * ==========================(LICENSE BEGIN)============================
 *
 * Copyright (c) 2014  Savale
 *
 * Permission is hereby granted, free of charge, to any person obtaining
 * a copy of this software and associated documentation files (the
 * "Software"), to deal in the Software without restriction, including
 * without limitation the rights to use, copy, modify, merge, publish,
 * distribute, sublicense, and/or sell copies of the Software, and to
 * permit persons to whom the Software is furnished to do so, subject to
 * the following conditions:
 *
 * The above copyright notice and this permission notice shall be
 * included in all copies or substantial portions of the Software.
 *
 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
 * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
 * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
 * IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
 * CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
 * TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
 * SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
 *
 * ===========================(LICENSE END)=============================
 *
 * @author   Savale <info@freshcoin.net>
 */

#ifndef FRESH_CL
#define FRESH_CL

#if __ENDIAN_LITTLE__
  #define SPH_LITTLE_ENDIAN 1
#else
  #define SPH_BIG_ENDIAN 1
#endif

#define SPH_UPTR sph_u64

typedef unsigned int sph_u32;
typedef int sph_s32;
#ifndef __OPENCL_VERSION__
  typedef unsigned long long sph_u64;
  typedef long long sph_s64;
#else
  typedef unsigned long sph_u64;
  typedef long sph_s64;
#endif

#define SPH_64 1
#define SPH_64_TRUE 1

#define SPH_C32(x) ((sph_u32)(x ## U))
#define SPH_T32(x) (as_uint(x))
#define SPH_ROTL32(x, n) rotate(as_uint(x), as_uint(n))
#define SPH_ROTR32(x, n) SPH_ROTL32(x, (32 - (n)))

#define SPH_C64(x) ((sph_u64)(x ## UL))
#define SPH_T64(x) (as_ulong(x))
#define SPH_ROTL64(x, n) rotate(as_ulong(x), (n) & 0xFFFFFFFFFFFFFFFFUL)
#define SPH_ROTR64(x, n) SPH_ROTL64(x, (64 - (n)))

#define SPH_ECHO_64 1
#define SPH_SIMD_NOCOPY 0

#include "shavite.cl"
#include "simd.cl"
#include "echo.cl"

#define SWAP4(x) as_uint(as_uchar4(x).wzyx)
#define SWAP8(x) as_ulong(as_uchar8(x).s76543210)

#if SPH_BIG_ENDIAN
  #define DEC64E(x) (x)
  #define DEC64BE(x) (*(const __global sph_u64 *) (x));
  #define DEC32LE(x) SWAP4(*(const __global sph_u32 *) (x));
#else
  #define DEC64E(x) SWAP8(x)
  #define DEC64BE(x) SWAP8(*(const __global sph_u64 *) (x));
  #define DEC32LE(x) (*(const __global sph_u32 *) (x));
#endif

typedef union {
  unsigned char h1[64];
  uint h4[16];
  ulong h8[8];
} hash_t;

__attribute__((reqd_work_group_size(WORKSIZE, 1, 1)))
__kernel void search(__global unsigned char* block, __global hash_t* hashes)
{
  uint gid = get_global_id(0);
  __global hash_t *hash = &(hashes[gid-get_global_offset(0)]);

  __local sph_u32 AES0[256], AES1[256], AES2[256], AES3[256];

  int init = get_local_id(0);
  int step = get_local_size(0);

  for (int i = init; i < 256; i += step)
  {
    AES0[i] = AES0_C[i];
    AES1[i] = AES1_C[i];
    AES2[i] = AES2_C[i];
    AES3[i] = AES3_C[i];
  }

  barrier(CLK_LOCAL_MEM_FENCE);

  // shavite
  // IV
  sph_u32 h0 = SPH_C32(0x72FCCDD8), h1 = SPH_C32(0x79CA4727), h2 = SPH_C32(0x128A077B), h3 = SPH_C32(0x40D55AEC);
  sph_u32 h4 = SPH_C32(0xD1901A06), h5 = SPH_C32(0x430AE307), h6 = SPH_C32(0xB29F5CD1), h7 = SPH_C32(0xDF07FBFC);
  sph_u32 h8 = SPH_C32(0x8E45D73D), h9 = SPH_C32(0x681AB538), hA = SPH_C32(0xBDE86578), hB = SPH_C32(0xDD577E47);
  sph_u32 hC = SPH_C32(0xE275EADE), hD = SPH_C32(0x502D9FCD), hE = SPH_C32(0xB9357178), hF = SPH_C32(0x022A4B9A);

  // state
  sph_u32 rk00, rk01, rk02, rk03, rk04, rk05, rk06, rk07;
  sph_u32 rk08, rk09, rk0A, rk0B, rk0C, rk0D, rk0E, rk0F;
  sph_u32 rk10, rk11, rk12, rk13, rk14, rk15, rk16, rk17;
  sph_u32 rk18, rk19, rk1A, rk1B, rk1C, rk1D, rk1E, rk1F;

  sph_u32 sc_count0 = 0x280, sc_count1 = 0, sc_count2 = 0, sc_count3 = 0;

  rk00 = DEC32LE(block + 0);
  rk01 = DEC32LE(block + 4);
  rk02 = DEC32LE(block + 8);
  rk03 = DEC32LE(block + 12);
  rk04 = DEC32LE(block + 16);
  rk05 = DEC32LE(block + 20);
  rk06 = DEC32LE(block + 24);
  rk07 = DEC32LE(block + 28);
  rk08 = DEC32LE(block + 32);
  rk09 = DEC32LE(block + 36);
  rk0A = DEC32LE(block + 40);
  rk0B = DEC32LE(block + 44);
  rk0C = DEC32LE(block + 48);
  rk0D = DEC32LE(block + 52);
  rk0E = DEC32LE(block + 56);
  rk0F = DEC32LE(block + 60);
  rk10 = DEC32LE(block + 64);
  rk11 = DEC32LE(block + 68);
  rk12 = DEC32LE(block + 72);
  rk13 = gid;
  rk14 = 0x80;
  rk15 = rk16 = rk17 = rk18 = rk19 = rk1A = 0;
  rk1B = 0x2800000;
  rk1C = rk1D = rk1E = 0;
  rk1F = 0x2000000;

  c512(buf);

  hash->h4[0] = h0;
  hash->h4[1] = h1;
  hash->h4[2] = h2;
  hash->h4[3] = h3;
  hash->h4[4] = h4;
  hash->h4[5] = h5;
  hash->h4[6] = h6;
  hash->h4[7] = h7;
  hash->h4[8] = h8;
  hash->h4[9] = h9;
  hash->h4[10] = hA;
  hash->h4[11] = hB;
  hash->h4[12] = hC;
  hash->h4[13] = hD;
  hash->h4[14] = hE;
  hash->h4[15] = hF;

  barrier(CLK_GLOBAL_MEM_FENCE);
}

__attribute__((reqd_work_group_size(WORKSIZE, 1, 1)))
__kernel void search1(__global hash_t* hashes)
{
  uint gid = get_global_id(0);
  __global hash_t *hash = &(hashes[gid-get_global_offset(0)]);

  // simd
  s32 q[256];
  unsigned char x[128];
  for(unsigned int i = 0; i < 64; i++)
    x[i] = hash->h1[i];
  for(unsigned int i = 64; i < 128; i++)
    x[i] = 0;

  u32 A0 = C32(0x0BA16B95), A1 = C32(0x72F999AD), A2 = C32(0x9FECC2AE), A3 = C32(0xBA3264FC), A4 = C32(0x5E894929), A5 = C32(0x8E9F30E5), A6 = C32(0x2F1DAA37), A7 = C32(0xF0F2C558);
  u32 B0 = C32(0xAC506643), B1 = C32(0xA90635A5), B2 = C32(0xE25B878B), B3 = C32(0xAAB7878F), B4 = C32(0x88817F7A), B5 = C32(0x0A02892B), B6 = C32(0x559A7550), B7 = C32(0x598F657E);
  u32 C0 = C32(0x7EEF60A1), C1 = C32(0x6B70E3E8), C2 = C32(0x9C1714D1), C3 = C32(0xB958E2A8), C4 = C32(0xAB02675E), C5 = C32(0xED1C014F), C6 = C32(0xCD8D65BB), C7 = C32(0xFDB7A257);
  u32 D0 = C32(0x09254899), D1 = C32(0xD699C7BC), D2 = C32(0x9019B6DC), D3 = C32(0x2B9022E4), D4 = C32(0x8FA14956), D5 = C32(0x21BF9BD3), D6 = C32(0xB94D0943), D7 = C32(0x6FFDDC22);

  FFT256(0, 1, 0, ll1);
  for (int i = 0; i < 256; i ++)
  {
    s32 tq;

    tq = q[i] + yoff_b_n[i];
    tq = REDS2(tq);
    tq = REDS1(tq);
    tq = REDS1(tq);
    q[i] = (tq <= 128 ? tq : tq - 257);
  }

  A0 ^= hash->h4[0];
  A1 ^= hash->h4[1];
  A2 ^= hash->h4[2];
  A3 ^= hash->h4[3];
  A4 ^= hash->h4[4];
  A5 ^= hash->h4[5];
  A6 ^= hash->h4[6];
  A7 ^= hash->h4[7];
  B0 ^= hash->h4[8];
  B1 ^= hash->h4[9];
  B2 ^= hash->h4[10];
  B3 ^= hash->h4[11];
  B4 ^= hash->h4[12];
  B5 ^= hash->h4[13];
  B6 ^= hash->h4[14];
  B7 ^= hash->h4[15];

  ONE_ROUND_BIG(0_, 0,  3, 23, 17, 27);
  ONE_ROUND_BIG(1_, 1, 28, 19, 22,  7);
  ONE_ROUND_BIG(2_, 2, 29,  9, 15,  5);
  ONE_ROUND_BIG(3_, 3,  4, 13, 10, 25);

  STEP_BIG(
    C32(0x0BA16B95), C32(0x72F999AD), C32(0x9FECC2AE), C32(0xBA3264FC),
    C32(0x5E894929), C32(0x8E9F30E5), C32(0x2F1DAA37), C32(0xF0F2C558),
    IF,  4, 13, PP8_4_);

  STEP_BIG(
    C32(0xAC506643), C32(0xA90635A5), C32(0xE25B878B), C32(0xAAB7878F),
    C32(0x88817F7A), C32(0x0A02892B), C32(0x559A7550), C32(0x598F657E),
    IF, 13, 10, PP8_5_);

  STEP_BIG(
    C32(0x7EEF60A1), C32(0x6B70E3E8), C32(0x9C1714D1), C32(0xB958E2A8),
    C32(0xAB02675E), C32(0xED1C014F), C32(0xCD8D65BB), C32(0xFDB7A257),
    IF, 10, 25, PP8_6_);

  STEP_BIG(
    C32(0x09254899), C32(0xD699C7BC), C32(0x9019B6DC), C32(0x2B9022E4),
    C32(0x8FA14956), C32(0x21BF9BD3), C32(0xB94D0943), C32(0x6FFDDC22),
    IF, 25,  4, PP8_0_);

  u32 COPY_A0 = A0, COPY_A1 = A1, COPY_A2 = A2, COPY_A3 = A3, COPY_A4 = A4, COPY_A5 = A5, COPY_A6 = A6, COPY_A7 = A7;
  u32 COPY_B0 = B0, COPY_B1 = B1, COPY_B2 = B2, COPY_B3 = B3, COPY_B4 = B4, COPY_B5 = B5, COPY_B6 = B6, COPY_B7 = B7;
  u32 COPY_C0 = C0, COPY_C1 = C1, COPY_C2 = C2, COPY_C3 = C3, COPY_C4 = C4, COPY_C5 = C5, COPY_C6 = C6, COPY_C7 = C7;
  u32 COPY_D0 = D0, COPY_D1 = D1, COPY_D2 = D2, COPY_D3 = D3, COPY_D4 = D4, COPY_D5 = D5, COPY_D6 = D6, COPY_D7 = D7;

  #define q SIMD_Q

  A0 ^= 0x200;

  ONE_ROUND_BIG(0_, 0,  3, 23, 17, 27);
  ONE_ROUND_BIG(1_, 1, 28, 19, 22,  7);
  ONE_ROUND_BIG(2_, 2, 29,  9, 15,  5);
  ONE_ROUND_BIG(3_, 3,  4, 13, 10, 25);

  STEP_BIG(
    COPY_A0, COPY_A1, COPY_A2, COPY_A3,
    COPY_A4, COPY_A5, COPY_A6, COPY_A7,
    IF,  4, 13, PP8_4_);

  STEP_BIG(
    COPY_B0, COPY_B1, COPY_B2, COPY_B3,
    COPY_B4, COPY_B5, COPY_B6, COPY_B7,
    IF, 13, 10, PP8_5_);

  STEP_BIG(
    COPY_C0, COPY_C1, COPY_C2, COPY_C3,
    COPY_C4, COPY_C5, COPY_C6, COPY_C7,
    IF, 10, 25, PP8_6_);

  STEP_BIG(
    COPY_D0, COPY_D1, COPY_D2, COPY_D3,
    COPY_D4, COPY_D5, COPY_D6, COPY_D7,
    IF, 25,  4, PP8_0_);

  #undef q

  hash->h4[0] = A0;
  hash->h4[1] = A1;
  hash->h4[2] = A2;
  hash->h4[3] = A3;
  hash->h4[4] = A4;
  hash->h4[5] = A5;
  hash->h4[6] = A6;
  hash->h4[7] = A7;
  hash->h4[8] = B0;
  hash->h4[9] = B1;
  hash->h4[10] = B2;
  hash->h4[11] = B3;
  hash->h4[12] = B4;
  hash->h4[13] = B5;
  hash->h4[14] = B6;
  hash->h4[15] = B7;

  barrier(CLK_GLOBAL_MEM_FENCE);
}

__attribute__((reqd_work_group_size(WORKSIZE, 1, 1)))
__kernel void search2(__global hash_t* hashes)
{
  uint gid = get_global_id(0);
  __global hash_t *hash = &(hashes[gid-get_global_offset(0)]);

  __local sph_u32 AES0[256], AES1[256], AES2[256], AES3[256];

  int init = get_local_id(0);
  int step = get_local_size(0);

  for (int i = init; i < 256; i += step)
  {
    AES0[i] = AES0_C[i];
    AES1[i] = AES1_C[i];
    AES2[i] = AES2_C[i];
    AES3[i] = AES3_C[i];
  }

  barrier(CLK_LOCAL_MEM_FENCE);

  // shavite
  // IV
  sph_u32 h0 = SPH_C32(0x72FCCDD8), h1 = SPH_C32(0x79CA4727), h2 = SPH_C32(0x128A077B), h3 = SPH_C32(0x40D55AEC);
  sph_u32 h4 = SPH_C32(0xD1901A06), h5 = SPH_C32(0x430AE307), h6 = SPH_C32(0xB29F5CD1), h7 = SPH_C32(0xDF07FBFC);
  sph_u32 h8 = SPH_C32(0x8E45D73D), h9 = SPH_C32(0x681AB538), hA = SPH_C32(0xBDE86578), hB = SPH_C32(0xDD577E47);
  sph_u32 hC = SPH_C32(0xE275EADE), hD = SPH_C32(0x502D9FCD), hE = SPH_C32(0xB9357178), hF = SPH_C32(0x022A4B9A);

  // state
  sph_u32 rk00, rk01, rk02, rk03, rk04, rk05, rk06, rk07;
  sph_u32 rk08, rk09, rk0A, rk0B, rk0C, rk0D, rk0E, rk0F;
  sph_u32 rk10, rk11, rk12, rk13, rk14, rk15, rk16, rk17;
  sph_u32 rk18, rk19, rk1A, rk1B, rk1C, rk1D, rk1E, rk1F;

  sph_u32 sc_count0 = 0x200, sc_count1 = 0, sc_count2 = 0, sc_count3 = 0;

  rk00 = hash->h4[0];
  rk01 = hash->h4[1];
  rk02 = hash->h4[2];
  rk03 = hash->h4[3];
  rk04 = hash->h4[4];
  rk05 = hash->h4[5];
  rk06 = hash->h4[6];
  rk07 = hash->h4[7];
  rk08 = hash->h4[8];
  rk09 = hash->h4[9];
  rk0A = hash->h4[10];
  rk0B = hash->h4[11];
  rk0C = hash->h4[12];
  rk0D = hash->h4[13];
  rk0E = hash->h4[14];
  rk0F = hash->h4[15];
  rk10 = 0x80;
  rk11 = rk12 = rk13 = rk14 = rk15 = rk16 = rk17 = rk18 = rk19 = rk1A = 0;
  rk1B = 0x2000000;
  rk1C = rk1D = rk1E = 0;
  rk1F = 0x2000000;

  c512(buf);

  hash->h4[0] = h0;
  hash->h4[1] = h1;
  hash->h4[2] = h2;
  hash->h4[3] = h3;
  hash->h4[4] = h4;
  hash->h4[5] = h5;
  hash->h4[6] = h6;
  hash->h4[7] = h7;
  hash->h4[8] = h8;
  hash->h4[9] = h9;
  hash->h4[10] = hA;
  hash->h4[11] = hB;
  hash->h4[12] = hC;
  hash->h4[13] = hD;
  hash->h4[14] = hE;
  hash->h4[15] = hF;

  barrier(CLK_GLOBAL_MEM_FENCE);
}

__attribute__((reqd_work_group_size(WORKSIZE, 1, 1)))
__kernel void search3(__global hash_t* hashes)
{
  uint gid = get_global_id(0);
  __global hash_t *hash = &(hashes[gid-get_global_offset(0)]);

  // simd
  s32 q[256];
  unsigned char x[128];
  for(unsigned int i = 0; i < 64; i++)
    x[i] = hash->h1[i];
  for(unsigned int i = 64; i < 128; i++)
    x[i] = 0;

  u32 A0 = C32(0x0BA16B95), A1 = C32(0x72F999AD), A2 = C32(0x9FECC2AE), A3 = C32(0xBA3264FC), A4 = C32(0x5E894929), A5 = C32(0x8E9F30E5), A6 = C32(0x2F1DAA37), A7 = C32(0xF0F2C558);
  u32 B0 = C32(0xAC506643), B1 = C32(0xA90635A5), B2 = C32(0xE25B878B), B3 = C32(0xAAB7878F), B4 = C32(0x88817F7A), B5 = C32(0x0A02892B), B6 = C32(0x559A7550), B7 = C32(0x598F657E);
  u32 C0 = C32(0x7EEF60A1), C1 = C32(0x6B70E3E8), C2 = C32(0x9C1714D1), C3 = C32(0xB958E2A8), C4 = C32(0xAB02675E), C5 = C32(0xED1C014F), C6 = C32(0xCD8D65BB), C7 = C32(0xFDB7A257);
  u32 D0 = C32(0x09254899), D1 = C32(0xD699C7BC), D2 = C32(0x9019B6DC), D3 = C32(0x2B9022E4), D4 = C32(0x8FA14956), D5 = C32(0x21BF9BD3), D6 = C32(0xB94D0943), D7 = C32(0x6FFDDC22);

  FFT256(0, 1, 0, ll1);
  for (int i = 0; i < 256; i ++)
  {
    s32 tq;

    tq = q[i] + yoff_b_n[i];
    tq = REDS2(tq);
    tq = REDS1(tq);
    tq = REDS1(tq);
    q[i] = (tq <= 128 ? tq : tq - 257);
  }

  A0 ^= hash->h4[0];
  A1 ^= hash->h4[1];
  A2 ^= hash->h4[2];
  A3 ^= hash->h4[3];
  A4 ^= hash->h4[4];
  A5 ^= hash->h4[5];
  A6 ^= hash->h4[6];
  A7 ^= hash->h4[7];
  B0 ^= hash->h4[8];
  B1 ^= hash->h4[9];
  B2 ^= hash->h4[10];
  B3 ^= hash->h4[11];
  B4 ^= hash->h4[12];
  B5 ^= hash->h4[13];
  B6 ^= hash->h4[14];
  B7 ^= hash->h4[15];

  ONE_ROUND_BIG(0_, 0,  3, 23, 17, 27);
  ONE_ROUND_BIG(1_, 1, 28, 19, 22,  7);
  ONE_ROUND_BIG(2_, 2, 29,  9, 15,  5);
  ONE_ROUND_BIG(3_, 3,  4, 13, 10, 25);

  STEP_BIG(
    C32(0x0BA16B95), C32(0x72F999AD), C32(0x9FECC2AE), C32(0xBA3264FC),
    C32(0x5E894929), C32(0x8E9F30E5), C32(0x2F1DAA37), C32(0xF0F2C558),
    IF,  4, 13, PP8_4_);

  STEP_BIG(
    C32(0xAC506643), C32(0xA90635A5), C32(0xE25B878B), C32(0xAAB7878F),
    C32(0x88817F7A), C32(0x0A02892B), C32(0x559A7550), C32(0x598F657E),
    IF, 13, 10, PP8_5_);

  STEP_BIG(
    C32(0x7EEF60A1), C32(0x6B70E3E8), C32(0x9C1714D1), C32(0xB958E2A8),
    C32(0xAB02675E), C32(0xED1C014F), C32(0xCD8D65BB), C32(0xFDB7A257),
    IF, 10, 25, PP8_6_);

  STEP_BIG(
    C32(0x09254899), C32(0xD699C7BC), C32(0x9019B6DC), C32(0x2B9022E4),
    C32(0x8FA14956), C32(0x21BF9BD3), C32(0xB94D0943), C32(0x6FFDDC22),
    IF, 25,  4, PP8_0_);

  u32 COPY_A0 = A0, COPY_A1 = A1, COPY_A2 = A2, COPY_A3 = A3, COPY_A4 = A4, COPY_A5 = A5, COPY_A6 = A6, COPY_A7 = A7;
  u32 COPY_B0 = B0, COPY_B1 = B1, COPY_B2 = B2, COPY_B3 = B3, COPY_B4 = B4, COPY_B5 = B5, COPY_B6 = B6, COPY_B7 = B7;
  u32 COPY_C0 = C0, COPY_C1 = C1, COPY_C2 = C2, COPY_C3 = C3, COPY_C4 = C4, COPY_C5 = C5, COPY_C6 = C6, COPY_C7 = C7;
  u32 COPY_D0 = D0, COPY_D1 = D1, COPY_D2 = D2, COPY_D3 = D3, COPY_D4 = D4, COPY_D5 = D5, COPY_D6 = D6, COPY_D7 = D7;

  #define q SIMD_Q

  A0 ^= 0x200;

  ONE_ROUND_BIG(0_, 0,  3, 23, 17, 27);
  ONE_ROUND_BIG(1_, 1, 28, 19, 22,  7);
  ONE_ROUND_BIG(2_, 2, 29,  9, 15,  5);
  ONE_ROUND_BIG(3_, 3,  4, 13, 10, 25);

  STEP_BIG(
    COPY_A0, COPY_A1, COPY_A2, COPY_A3,
    COPY_A4, COPY_A5, COPY_A6, COPY_A7,
    IF,  4, 13, PP8_4_);

  STEP_BIG(
    COPY_B0, COPY_B1, COPY_B2, COPY_B3,
    COPY_B4, COPY_B5, COPY_B6, COPY_B7,
    IF, 13, 10, PP8_5_);

  STEP_BIG(
    COPY_C0, COPY_C1, COPY_C2, COPY_C3,
    COPY_C4, COPY_C5, COPY_C6, COPY_C7,
    IF, 10, 25, PP8_6_);

  STEP_BIG(
    COPY_D0, COPY_D1, COPY_D2, COPY_D3,
    COPY_D4, COPY_D5, COPY_D6, COPY_D7,
    IF, 25,  4, PP8_0_);

  #undef q

  hash->h4[0] = A0;
  hash->h4[1] = A1;
  hash->h4[2] = A2;
  hash->h4[3] = A3;
  hash->h4[4] = A4;
  hash->h4[5] = A5;
  hash->h4[6] = A6;
  hash->h4[7] = A7;
  hash->h4[8] = B0;
  hash->h4[9] = B1;
  hash->h4[10] = B2;
  hash->h4[11] = B3;
  hash->h4[12] = B4;
  hash->h4[13] = B5;
  hash->h4[14] = B6;
  hash->h4[15] = B7;

  barrier(CLK_GLOBAL_MEM_FENCE);
}

__attribute__((reqd_work_group_size(WORKSIZE, 1, 1)))
__kernel void search4(__global hash_t* hashes, __global uint* output, const ulong target)
{
  uint gid = get_global_id(0);
  __global hash_t *hash = &(hashes[gid-get_global_offset(0)]);

  __local sph_u32 AES0[256], AES1[256], AES2[256], AES3[256];

  int init = get_local_id(0);
  int step = get_local_size(0);

  for (int i = init; i < 256; i += step)
  {
    AES0[i] = AES0_C[i];
    AES1[i] = AES1_C[i];
    AES2[i] = AES2_C[i];
    AES3[i] = AES3_C[i];
  }

  barrier(CLK_LOCAL_MEM_FENCE);

  // echo
  sph_u64 W00, W01, W10, W11, W20, W21, W30, W31, W40, W41, W50, W51, W60, W61, W70, W71, W80, W81, W90, W91, WA0, WA1, WB0, WB1, WC0, WC1, WD0, WD1, WE0, WE1, WF0, WF1;
  sph_u64 Vb00, Vb01, Vb10, Vb11, Vb20, Vb21, Vb30, Vb31, Vb40, Vb41, Vb50, Vb51, Vb60, Vb61, Vb70, Vb71;
  Vb00 = Vb10 = Vb20 = Vb30 = Vb40 = Vb50 = Vb60 = Vb70 = 512UL;
  Vb01 = Vb11 = Vb21 = Vb31 = Vb41 = Vb51 = Vb61 = Vb71 = 0;

  sph_u32 K0 = 512;
  sph_u32 K1 = 0;
  sph_u32 K2 = 0;
  sph_u32 K3 = 0;

  W00 = Vb00;
  W01 = Vb01;
  W10 = Vb10;
  W11 = Vb11;
  W20 = Vb20;
  W21 = Vb21;
  W30 = Vb30;
  W31 = Vb31;
  W40 = Vb40;
  W41 = Vb41;
  W50 = Vb50;
  W51 = Vb51;
  W60 = Vb60;
  W61 = Vb61;
  W70 = Vb70;
  W71 = Vb71;
  W80 = hash->h8[0];
  W81 = hash->h8[1];
  W90 = hash->h8[2];
  W91 = hash->h8[3];
  WA0 = hash->h8[4];
  WA1 = hash->h8[5];
  WB0 = hash->h8[6];
  WB1 = hash->h8[7];
  WC0 = 0x80;
  WC1 = 0;
  WD0 = 0;
  WD1 = 0;
  WE0 = 0;
  WE1 = 0x200000000000000;
  WF0 = 0x200;
  WF1 = 0;

  for (unsigned u = 0; u < 10; u ++)
    BIG_ROUND;

  Vb00 ^= hash->h8[0] ^ W00 ^ W80;
  Vb01 ^= hash->h8[1] ^ W01 ^ W81;
  Vb10 ^= hash->h8[2] ^ W10 ^ W90;
  Vb11 ^= hash->h8[3] ^ W11 ^ W91;
  Vb20 ^= hash->h8[4] ^ W20 ^ WA0;
  Vb21 ^= hash->h8[5] ^ W21 ^ WA1;
  Vb30 ^= hash->h8[6] ^ W30 ^ WB0;
  Vb31 ^= hash->h8[7] ^ W31 ^ WB1;

  bool result = (Vb11 <= target);
  if (result)
      output[output[0xFF]++] = SWAP4(gid);
}

#endif // FRESH_CL