/**
 * Author......: Jens Steube <jens.steube@gmail.com>
 * License.....: MIT
 */

#define _MD5_

#include "include/constants.h"
#include "include/kernel_vendor.h"

#ifdef  VLIW1
#define VECT_SIZE1
#endif

#ifdef  VLIW4
#define VECT_SIZE1
#endif

#ifdef  VLIW5
#define VECT_SIZE1
#endif

#define DGST_R0 0
#define DGST_R1 1
#define DGST_R2 2
#define DGST_R3 3

#include "include/kernel_functions.c"
#include "types_amd.c"
#include "common_amd.c"
#include "include/rp_gpu.h"
#include "rp_amd.c"

#ifdef  VECT_SIZE1
#define VECT_COMPARE_S "check_single_vect1_comp4.c"
#define VECT_COMPARE_M "check_multi_vect1_comp4.c"
#endif

#ifdef  VECT_SIZE4
#define VECT_COMPARE_S "check_single_vect4_comp4.c"
#define VECT_COMPARE_M "check_multi_vect4_comp4.c"
#endif

__constant u32 padding[8] =
{
  0x5e4ebf28,
  0x418a754e,
  0x564e0064,
  0x0801faff,
  0xb6002e2e,
  0x803e68d0,
  0xfea90c2f,
  0x7a695364
};

typedef struct
{
  u8 S[256];

  u32 wtf_its_faster;

} RC4_KEY;

static void swap (__local RC4_KEY *rc4_key, const u8 i, const u8 j)
{
  u8 tmp;

  tmp           = rc4_key->S[i];
  rc4_key->S[i] = rc4_key->S[j];
  rc4_key->S[j] = tmp;
}

static void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
{
  u32 v = 0x03020100;
  u32 a = 0x04040404;

  __local u32 *ptr = (__local u32 *) rc4_key->S;

  #pragma unroll
  for (u32 i = 0; i < 64; i++)
  {
    ptr[i] = v; v += a;
  }

  const u32 d0 = data[0] >>  0;
  const u32 d1 = data[0] >>  8;
  const u32 d2 = data[0] >> 16;
  const u32 d3 = data[0] >> 24;
  const u32 d4 = data[1] >>  0;

  u32 j = 0;

  #pragma unroll
  for (u32 i = 0; i < 255; i += 5)
  {
    j += rc4_key->S[i + 0] + d0; swap (rc4_key, i + 0, j);
    j += rc4_key->S[i + 1] + d1; swap (rc4_key, i + 1, j);
    j += rc4_key->S[i + 2] + d2; swap (rc4_key, i + 2, j);
    j += rc4_key->S[i + 3] + d3; swap (rc4_key, i + 3, j);
    j += rc4_key->S[i + 4] + d4; swap (rc4_key, i + 4, j);
  }

  j += rc4_key->S[255] + d0; swap (rc4_key, 255, j);
}

static u8 rc4_next_16 (__local RC4_KEY *rc4_key, u8 i, u8 j, __constant u32 in[4], u32 out[4])
{
  for (u32 k = 0; k < 4; k++)
  {
    u32 xor4 = 0;

    u8 idx;

    i += 1;
    j += rc4_key->S[i];

    swap (rc4_key, i, j);

    idx = rc4_key->S[i] + rc4_key->S[j];

    xor4 |= rc4_key->S[idx] <<  0;

    i += 1;
    j += rc4_key->S[i];

    swap (rc4_key, i, j);

    idx = rc4_key->S[i] + rc4_key->S[j];

    xor4 |= rc4_key->S[idx] <<  8;

    i += 1;
    j += rc4_key->S[i];

    swap (rc4_key, i, j);

    idx = rc4_key->S[i] + rc4_key->S[j];

    xor4 |= rc4_key->S[idx] << 16;

    i += 1;
    j += rc4_key->S[i];

    swap (rc4_key, i, j);

    idx = rc4_key->S[i] + rc4_key->S[j];

    xor4 |= rc4_key->S[idx] << 24;

    out[k] = in[k] ^ xor4;
  }

  return j;
}

__kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m10410_m04 (__global pw_t *pws, __global gpu_rule_t *rules_buf, __global comb_t *combs_buf, __global bf_t *bfs_buf, __global void *tmps, __global void *hooks, __global u32 *bitmaps_buf_s1_a, __global u32 *bitmaps_buf_s1_b, __global u32 *bitmaps_buf_s1_c, __global u32 *bitmaps_buf_s1_d, __global u32 *bitmaps_buf_s2_a, __global u32 *bitmaps_buf_s2_b, __global u32 *bitmaps_buf_s2_c, __global u32 *bitmaps_buf_s2_d, __global plain_t *plains_buf, __global digest_t *digests_buf, __global u32 *hashes_shown, __global salt_t *salt_bufs, __global pdf_t *pdf_bufs, __global u32 *d_return_buf, __global u32 *d_scryptV_buf, const u32 bitmap_mask, const u32 bitmap_shift1, const u32 bitmap_shift2, const u32 salt_pos, const u32 loop_pos, const u32 loop_cnt, const u32 rules_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
{
  /**
   * modifier
   */

  const u32 lid = get_local_id (0);

  /**
   * base
   */

  const u32 gid = get_global_id (0);

  if (gid >= gid_max) return;

  u32x pw_buf0[4];

  pw_buf0[0] = pws[gid].i[ 0];
  pw_buf0[1] = pws[gid].i[ 1];
  pw_buf0[2] = pws[gid].i[ 2];
  pw_buf0[3] = pws[gid].i[ 3];

  u32x pw_buf1[4];

  pw_buf1[0] = pws[gid].i[ 4];
  pw_buf1[1] = pws[gid].i[ 5];
  pw_buf1[2] = pws[gid].i[ 6];
  pw_buf1[3] = pws[gid].i[ 7];

  const u32 pw_len = pws[gid].pw_len;

  /**
   * key
   */

  __local RC4_KEY rc4_keys[64];
  __local RC4_KEY *rc4_key = &rc4_keys[lid];

  /**
   * loop
   */

  for (u32 il_pos = 0; il_pos < rules_cnt; il_pos++)
  {
    u32x w0[4];

    w0[0] = pw_buf0[0];
    w0[1] = pw_buf0[1];
    w0[2] = pw_buf0[2];
    w0[3] = pw_buf0[3];

    u32x w1[4];

    w1[0] = pw_buf1[0];
    w1[1] = pw_buf1[1];
    w1[2] = pw_buf1[2];
    w1[3] = pw_buf1[3];

    u32x w2[4];

    w2[0] = 0;
    w2[1] = 0;
    w2[2] = 0;
    w2[3] = 0;

    u32x w3[4];

    w3[0] = 0;
    w3[1] = 0;
    w3[2] = 0;
    w3[3] = 0;

    const u32 out_len = apply_rules (rules_buf[il_pos].cmds, w0, w1, pw_len);

    // now the RC4 part

    u32x key[4];

    key[0] = w0[0];
    key[1] = w0[1];
    key[2] = 0;
    key[3] = 0;

    rc4_init_16 (rc4_key, key);

    u32x out[4];

    rc4_next_16 (rc4_key, 0, 0, padding, out);

    const u32x r0 = out[0];
    const u32x r1 = out[1];
    const u32x r2 = out[2];
    const u32x r3 = out[3];

    #include VECT_COMPARE_M
  }
}

__kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m10410_m08 (__global pw_t *pws, __global gpu_rule_t *rules_buf, __global comb_t *combs_buf, __global bf_t *bfs_buf, __global void *tmps, __global void *hooks, __global u32 *bitmaps_buf_s1_a, __global u32 *bitmaps_buf_s1_b, __global u32 *bitmaps_buf_s1_c, __global u32 *bitmaps_buf_s1_d, __global u32 *bitmaps_buf_s2_a, __global u32 *bitmaps_buf_s2_b, __global u32 *bitmaps_buf_s2_c, __global u32 *bitmaps_buf_s2_d, __global plain_t *plains_buf, __global digest_t *digests_buf, __global u32 *hashes_shown, __global salt_t *salt_bufs, __global pdf_t *pdf_bufs, __global u32 *d_return_buf, __global u32 *d_scryptV_buf, const u32 bitmap_mask, const u32 bitmap_shift1, const u32 bitmap_shift2, const u32 salt_pos, const u32 loop_pos, const u32 loop_cnt, const u32 rules_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
{
}

__kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m10410_m16 (__global pw_t *pws, __global gpu_rule_t *rules_buf, __global comb_t *combs_buf, __global bf_t *bfs_buf, __global void *tmps, __global void *hooks, __global u32 *bitmaps_buf_s1_a, __global u32 *bitmaps_buf_s1_b, __global u32 *bitmaps_buf_s1_c, __global u32 *bitmaps_buf_s1_d, __global u32 *bitmaps_buf_s2_a, __global u32 *bitmaps_buf_s2_b, __global u32 *bitmaps_buf_s2_c, __global u32 *bitmaps_buf_s2_d, __global plain_t *plains_buf, __global digest_t *digests_buf, __global u32 *hashes_shown, __global salt_t *salt_bufs, __global pdf_t *pdf_bufs, __global u32 *d_return_buf, __global u32 *d_scryptV_buf, const u32 bitmap_mask, const u32 bitmap_shift1, const u32 bitmap_shift2, const u32 salt_pos, const u32 loop_pos, const u32 loop_cnt, const u32 rules_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
{
}

__kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m10410_s04 (__global pw_t *pws, __global gpu_rule_t *rules_buf, __global comb_t *combs_buf, __global bf_t *bfs_buf, __global void *tmps, __global void *hooks, __global u32 *bitmaps_buf_s1_a, __global u32 *bitmaps_buf_s1_b, __global u32 *bitmaps_buf_s1_c, __global u32 *bitmaps_buf_s1_d, __global u32 *bitmaps_buf_s2_a, __global u32 *bitmaps_buf_s2_b, __global u32 *bitmaps_buf_s2_c, __global u32 *bitmaps_buf_s2_d, __global plain_t *plains_buf, __global digest_t *digests_buf, __global u32 *hashes_shown, __global salt_t *salt_bufs, __global pdf_t *pdf_bufs, __global u32 *d_return_buf, __global u32 *d_scryptV_buf, const u32 bitmap_mask, const u32 bitmap_shift1, const u32 bitmap_shift2, const u32 salt_pos, const u32 loop_pos, const u32 loop_cnt, const u32 rules_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
{
  /**
   * modifier
   */

  const u32 lid = get_local_id (0);

  /**
   * base
   */

  const u32 gid = get_global_id (0);

  if (gid >= gid_max) return;

  u32x pw_buf0[4];

  pw_buf0[0] = pws[gid].i[ 0];
  pw_buf0[1] = pws[gid].i[ 1];
  pw_buf0[2] = pws[gid].i[ 2];
  pw_buf0[3] = pws[gid].i[ 3];

  u32x pw_buf1[4];

  pw_buf1[0] = pws[gid].i[ 4];
  pw_buf1[1] = pws[gid].i[ 5];
  pw_buf1[2] = pws[gid].i[ 6];
  pw_buf1[3] = pws[gid].i[ 7];

  const u32 pw_len = pws[gid].pw_len;

  /**
   * digest
   */

  const u32 search[4] =
  {
    digests_buf[digests_offset].digest_buf[DGST_R0],
    digests_buf[digests_offset].digest_buf[DGST_R1],
    digests_buf[digests_offset].digest_buf[DGST_R2],
    digests_buf[digests_offset].digest_buf[DGST_R3]
  };

  /**
   * key
   */

  __local RC4_KEY rc4_keys[64];
  __local RC4_KEY *rc4_key = &rc4_keys[lid];

  /**
   * loop
   */

  for (u32 il_pos = 0; il_pos < rules_cnt; il_pos++)
  {
    u32x w0[4];

    w0[0] = pw_buf0[0];
    w0[1] = pw_buf0[1];
    w0[2] = pw_buf0[2];
    w0[3] = pw_buf0[3];

    u32x w1[4];

    w1[0] = pw_buf1[0];
    w1[1] = pw_buf1[1];
    w1[2] = pw_buf1[2];
    w1[3] = pw_buf1[3];

    u32x w2[4];

    w2[0] = 0;
    w2[1] = 0;
    w2[2] = 0;
    w2[3] = 0;

    u32x w3[4];

    w3[0] = 0;
    w3[1] = 0;
    w3[2] = 0;
    w3[3] = 0;

    const u32 out_len = apply_rules (rules_buf[il_pos].cmds, w0, w1, pw_len);

    // now the RC4 part

    u32x key[4];

    key[0] = w0[0];
    key[1] = w0[1];
    key[2] = 0;
    key[3] = 0;

    rc4_init_16 (rc4_key, key);

    u32x out[4];

    rc4_next_16 (rc4_key, 0, 0, padding, out);

    const u32x r0 = out[0];
    const u32x r1 = out[1];
    const u32x r2 = out[2];
    const u32x r3 = out[3];

    #include VECT_COMPARE_S
  }
}

__kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m10410_s08 (__global pw_t *pws, __global gpu_rule_t *rules_buf, __global comb_t *combs_buf, __global bf_t *bfs_buf, __global void *tmps, __global void *hooks, __global u32 *bitmaps_buf_s1_a, __global u32 *bitmaps_buf_s1_b, __global u32 *bitmaps_buf_s1_c, __global u32 *bitmaps_buf_s1_d, __global u32 *bitmaps_buf_s2_a, __global u32 *bitmaps_buf_s2_b, __global u32 *bitmaps_buf_s2_c, __global u32 *bitmaps_buf_s2_d, __global plain_t *plains_buf, __global digest_t *digests_buf, __global u32 *hashes_shown, __global salt_t *salt_bufs, __global pdf_t *pdf_bufs, __global u32 *d_return_buf, __global u32 *d_scryptV_buf, const u32 bitmap_mask, const u32 bitmap_shift1, const u32 bitmap_shift2, const u32 salt_pos, const u32 loop_pos, const u32 loop_cnt, const u32 rules_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
{
}

__kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m10410_s16 (__global pw_t *pws, __global gpu_rule_t *rules_buf, __global comb_t *combs_buf, __global bf_t *bfs_buf, __global void *tmps, __global void *hooks, __global u32 *bitmaps_buf_s1_a, __global u32 *bitmaps_buf_s1_b, __global u32 *bitmaps_buf_s1_c, __global u32 *bitmaps_buf_s1_d, __global u32 *bitmaps_buf_s2_a, __global u32 *bitmaps_buf_s2_b, __global u32 *bitmaps_buf_s2_c, __global u32 *bitmaps_buf_s2_d, __global plain_t *plains_buf, __global digest_t *digests_buf, __global u32 *hashes_shown, __global salt_t *salt_bufs, __global pdf_t *pdf_bufs, __global u32 *d_return_buf, __global u32 *d_scryptV_buf, const u32 bitmap_mask, const u32 bitmap_shift1, const u32 bitmap_shift2, const u32 salt_pos, const u32 loop_pos, const u32 loop_cnt, const u32 rules_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
{
}