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

#define _SAPG_

//incompatible data-dependant code
//#define NEW_SIMD_CODE

#include "inc_vendor.cl"
#include "inc_hash_constants.h"
#include "inc_hash_functions.cl"
#include "inc_types.cl"
#include "inc_common.cl"
#include "inc_simd.cl"

#define GETSHIFTEDINT(a,n) amd_bytealign ((a)[((n)/4)+1], (a)[((n)/4)+0], (n))

#define SETSHIFTEDINT(a,n,v)        \
{                                   \
  const u32 s = ((n) & 3) * 8;     \
  const u64 x = (u64) (v) << s; \
  (a)[((n)/4)+0] |= x;              \
  (a)[((n)/4)+1]  = x >> 32;        \
}

__constant u32 theMagicArray[64] =
{
  0x1451ac91,0x4354679f,0xe03be724,0xc27b7428,0xeb133386,0x5ccb4f5a,0x37730a08,0x2f1c5d0e,
  0xe5e68f33,0xddae9bf8,0x8d4bf216,0xdcd4e12c,0x9ddfcbb0,0x176d70d4,0x3f424df9,0x94111b9b,
  0x9bc15b9f,0x039d0506,0x8a135e9d,0xe86a9a1e,0x17147cd9,0xf62ac758,0x0a6399a1,0xc370fdd7,
  0x13745ef6,0x040bc903,0x26f79826,0x2593928a,0x230da2b0,0x6d7963ed,0x3cfa3213,0xa39a0235,
  0x0a8eddb3,0xc351bf24,0x9f55cd7c,0x4c94af37,0x82520829,0x374e3bb2,0x9107179f,0xcdfd3b11,
  0, 0, 0, 0, 0, 0, 0, 0,
  0, 0, 0, 0, 0, 0, 0, 0,
  0, 0, 0, 0, 0, 0, 0, 0
};

void swap_buffer (u32 final[16])
{
  final[ 0] = swap32 (final[ 0]);
  final[ 1] = swap32 (final[ 1]);
  final[ 2] = swap32 (final[ 2]);
  final[ 3] = swap32 (final[ 3]);
  final[ 4] = swap32 (final[ 4]);
  final[ 5] = swap32 (final[ 5]);
  final[ 6] = swap32 (final[ 6]);
  final[ 7] = swap32 (final[ 7]);
  final[ 8] = swap32 (final[ 8]);
  final[ 9] = swap32 (final[ 9]);
  final[10] = swap32 (final[10]);
  final[11] = swap32 (final[11]);
  final[12] = swap32 (final[12]);
  final[13] = swap32 (final[13]);
  final[14] = swap32 (final[14]);
  final[15] = swap32 (final[15]);
}

void sha1_transform (const u32 w0[4], const u32 w1[4], const u32 w2[4], const u32 w3[4], u32 digest[5])
{
  u32 A = digest[0];
  u32 B = digest[1];
  u32 C = digest[2];
  u32 D = digest[3];
  u32 E = digest[4];

  u32 w0_t = w0[0];
  u32 w1_t = w0[1];
  u32 w2_t = w0[2];
  u32 w3_t = w0[3];
  u32 w4_t = w1[0];
  u32 w5_t = w1[1];
  u32 w6_t = w1[2];
  u32 w7_t = w1[3];
  u32 w8_t = w2[0];
  u32 w9_t = w2[1];
  u32 wa_t = w2[2];
  u32 wb_t = w2[3];
  u32 wc_t = w3[0];
  u32 wd_t = w3[1];
  u32 we_t = w3[2];
  u32 wf_t = w3[3];

  #undef K
  #define K SHA1C00

  SHA1_STEP (SHA1_F0o, A, B, C, D, E, w0_t);
  SHA1_STEP (SHA1_F0o, E, A, B, C, D, w1_t);
  SHA1_STEP (SHA1_F0o, D, E, A, B, C, w2_t);
  SHA1_STEP (SHA1_F0o, C, D, E, A, B, w3_t);
  SHA1_STEP (SHA1_F0o, B, C, D, E, A, w4_t);
  SHA1_STEP (SHA1_F0o, A, B, C, D, E, w5_t);
  SHA1_STEP (SHA1_F0o, E, A, B, C, D, w6_t);
  SHA1_STEP (SHA1_F0o, D, E, A, B, C, w7_t);
  SHA1_STEP (SHA1_F0o, C, D, E, A, B, w8_t);
  SHA1_STEP (SHA1_F0o, B, C, D, E, A, w9_t);
  SHA1_STEP (SHA1_F0o, A, B, C, D, E, wa_t);
  SHA1_STEP (SHA1_F0o, E, A, B, C, D, wb_t);
  SHA1_STEP (SHA1_F0o, D, E, A, B, C, wc_t);
  SHA1_STEP (SHA1_F0o, C, D, E, A, B, wd_t);
  SHA1_STEP (SHA1_F0o, B, C, D, E, A, we_t);
  SHA1_STEP (SHA1_F0o, A, B, C, D, E, wf_t);
  w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F0o, E, A, B, C, D, w0_t);
  w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F0o, D, E, A, B, C, w1_t);
  w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F0o, C, D, E, A, B, w2_t);
  w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F0o, B, C, D, E, A, w3_t);

  #undef K
  #define K SHA1C01

  w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w4_t);
  w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w5_t);
  w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w6_t);
  w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w7_t);
  w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w8_t);
  w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w9_t);
  wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wa_t);
  wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, wb_t);
  wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, wc_t);
  wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, wd_t);
  we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, we_t);
  wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wf_t);
  w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w0_t);
  w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w1_t);
  w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w2_t);
  w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w3_t);
  w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w4_t);
  w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w5_t);
  w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w6_t);
  w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w7_t);

  #undef K
  #define K SHA1C02

  w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, w8_t);
  w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, w9_t);
  wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, wa_t);
  wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, wb_t);
  wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, wc_t);
  wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, wd_t);
  we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, we_t);
  wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, wf_t);
  w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, w0_t);
  w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, w1_t);
  w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, w2_t);
  w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, w3_t);
  w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, w4_t);
  w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, w5_t);
  w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, w6_t);
  w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, w7_t);
  w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, w8_t);
  w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, w9_t);
  wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, wa_t);
  wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, wb_t);

  #undef K
  #define K SHA1C03

  wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, wc_t);
  wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wd_t);
  we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, we_t);
  wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, wf_t);
  w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w0_t);
  w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w1_t);
  w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w2_t);
  w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w3_t);
  w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w4_t);
  w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w5_t);
  w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w6_t);
  w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w7_t);
  w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w8_t);
  w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w9_t);
  wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, wa_t);
  wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, wb_t);
  wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wc_t);
  wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, wd_t);
  we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, we_t);
  wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, wf_t);

  digest[0] += A;
  digest[1] += B;
  digest[2] += C;
  digest[3] += D;
  digest[4] += E;
}

void m07800m (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], const u32 pw_len, __global pw_t *pws, __global kernel_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 void *esalt_bufs, __global u32 *d_return_buf, __global u32 *d_scryptV0_buf, __global u32 *d_scryptV1_buf, __global u32 *d_scryptV2_buf, __global u32 *d_scryptV3_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 il_cnt, const u32 digests_cnt, const u32 digests_offset)
{
  /**
   * modifier
   */

  const u32 gid = get_global_id (0);
  const u32 lid = get_local_id (0);

  /**
   * salt
   */

  u32 salt_buf[8];

  salt_buf[0] = salt_bufs[salt_pos].salt_buf[0];
  salt_buf[1] = salt_bufs[salt_pos].salt_buf[1];
  salt_buf[2] = salt_bufs[salt_pos].salt_buf[2];
  salt_buf[3] = salt_bufs[salt_pos].salt_buf[3];
  salt_buf[4] = salt_bufs[salt_pos].salt_buf[4];
  salt_buf[5] = salt_bufs[salt_pos].salt_buf[5];
  salt_buf[6] = salt_bufs[salt_pos].salt_buf[6];
  salt_buf[7] = salt_bufs[salt_pos].salt_buf[7];

  const u32 salt_len = salt_bufs[salt_pos].salt_len;

  u32 s0[4];
  u32 s1[4];
  u32 s2[4];
  u32 s3[4];

  s0[0] = swap32 (salt_buf[0]);
  s0[1] = swap32 (salt_buf[1]);
  s0[2] = swap32 (salt_buf[2]);
  s0[3] = swap32 (salt_buf[3]);
  s1[0] = swap32 (salt_buf[4]);
  s1[1] = swap32 (salt_buf[5]);
  s1[2] = swap32 (salt_buf[6]);
  s1[3] = swap32 (salt_buf[7]);
  s2[0] = 0;
  s2[1] = 0;
  s2[2] = 0;
  s2[3] = 0;
  s3[0] = 0;
  s3[1] = 0;
  s3[2] = 0;
  s3[3] = 0;

  switch_buffer_by_offset_be_S (s0, s1, s2, s3, pw_len);

  const u32x pw_salt_len = pw_len + salt_len;

  /**
   * loop
   */

  u32 w0l = w0[0];

  for (u32 il_pos = 0; il_pos < il_cnt; il_pos += VECT_SIZE)
  {
    const u32x w0r = ix_create_bft (bfs_buf, il_pos);

    const u32x w0lr = w0l | w0r;

    w0[0] = w0lr;

    /**
     * SAP
     */

    u32 final[32];

    final[ 0] = w0[0] | s0[0];
    final[ 1] = w0[1] | s0[1];
    final[ 2] = w0[2] | s0[2];
    final[ 3] = w0[3] | s0[3];
    final[ 4] = w1[0] | s1[0];
    final[ 5] = w1[1] | s1[1];
    final[ 6] = w1[2] | s1[2];
    final[ 7] = w1[3] | s1[3];
    final[ 8] = w2[0] | s2[0];
    final[ 9] = w2[1] | s2[1];
    final[10] = w2[2] | s2[2];
    final[11] = w2[3] | s2[3];
    final[12] = w3[0] | s3[0];
    final[13] = w3[1] | s3[1];
    final[14] = 0;
    final[15] = pw_salt_len * 8;

    u32 digest[5];

    digest[0] = SHA1M_A;
    digest[1] = SHA1M_B;
    digest[2] = SHA1M_C;
    digest[3] = SHA1M_D;
    digest[4] = SHA1M_E;

    sha1_transform (&final[0], &final[4], &final[8], &final[12], digest);

    // prepare magic array range

    u32 lengthMagicArray = 0x20;
    u32 offsetMagicArray = 0;

    lengthMagicArray += ((digest[0] >> 24) & 0xff) % 6;
    lengthMagicArray += ((digest[0] >> 16) & 0xff) % 6;
    lengthMagicArray += ((digest[0] >>  8) & 0xff) % 6;
    lengthMagicArray += ((digest[0] >>  0) & 0xff) % 6;
    lengthMagicArray += ((digest[1] >> 24) & 0xff) % 6;
    lengthMagicArray += ((digest[1] >> 16) & 0xff) % 6;
    lengthMagicArray += ((digest[1] >>  8) & 0xff) % 6;
    lengthMagicArray += ((digest[1] >>  0) & 0xff) % 6;
    lengthMagicArray += ((digest[2] >> 24) & 0xff) % 6;
    lengthMagicArray += ((digest[2] >> 16) & 0xff) % 6;
    offsetMagicArray += ((digest[2] >>  8) & 0xff) % 8;
    offsetMagicArray += ((digest[2] >>  0) & 0xff) % 8;
    offsetMagicArray += ((digest[3] >> 24) & 0xff) % 8;
    offsetMagicArray += ((digest[3] >> 16) & 0xff) % 8;
    offsetMagicArray += ((digest[3] >>  8) & 0xff) % 8;
    offsetMagicArray += ((digest[3] >>  0) & 0xff) % 8;
    offsetMagicArray += ((digest[4] >> 24) & 0xff) % 8;
    offsetMagicArray += ((digest[4] >> 16) & 0xff) % 8;
    offsetMagicArray += ((digest[4] >>  8) & 0xff) % 8;
    offsetMagicArray += ((digest[4] >>  0) & 0xff) % 8;

    // final

    digest[0] = SHA1M_A;
    digest[1] = SHA1M_B;
    digest[2] = SHA1M_C;
    digest[3] = SHA1M_D;
    digest[4] = SHA1M_E;

    #ifdef _unroll
    #pragma unroll
    #endif
    for (int i = 0; i < 32; i++) final[i] = 0;

    final[0] = swap32 (w0[0]);
    final[1] = swap32 (w0[1]);
    final[2] = swap32 (w0[2]);
    final[3] = swap32 (w0[3]);
    final[4] = swap32 (w1[0]);
    final[5] = swap32 (w1[1]);
    final[6] = swap32 (w1[2]);
    final[7] = swap32 (w1[3]);

    u32 final_len = pw_len;

    int i;

    // append MagicArray

    for (i = 0; i < lengthMagicArray - 4; i += 4)
    {
      const u32 tmp = GETSHIFTEDINT (theMagicArray, offsetMagicArray + i);

      SETSHIFTEDINT (final, final_len + i, tmp);
    }

    const u32 mask = 0xffffffff >> (((i - lengthMagicArray) & 3) * 8);

    const u32 tmp = GETSHIFTEDINT (theMagicArray, offsetMagicArray + i) & mask;

    SETSHIFTEDINT (final, final_len + i, tmp);

    final_len += lengthMagicArray;

    // append Salt

    for (i = 0; i < salt_len + 1; i += 4) // +1 for the 0x80
    {
      const u32 tmp = salt_buf[i / 4]; // attention, int[] not char[]

      SETSHIFTEDINT (final, final_len + i, tmp);
    }

    final_len += salt_len;

    // calculate

    int left;
    int off;

    for (left = final_len, off = 0; left >= 56; left -= 64, off += 16)
    {
      swap_buffer (&final[off]);

      sha1_transform (&final[off + 0], &final[off + 4], &final[off + 8], &final[off + 12], digest);
    }

    swap_buffer (&final[off]);

    final[off + 14] = 0;
    final[off + 15] = final_len * 8;

    sha1_transform (&final[off + 0], &final[off + 4], &final[off + 8], &final[off + 12], digest);

    COMPARE_M_SIMD (digest[3], digest[4], digest[2], digest[1]);
  }
}

void m07800s (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], const u32 pw_len, __global pw_t *pws, __global kernel_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 void *esalt_bufs, __global u32 *d_return_buf, __global u32 *d_scryptV0_buf, __global u32 *d_scryptV1_buf, __global u32 *d_scryptV2_buf, __global u32 *d_scryptV3_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 il_cnt, const u32 digests_cnt, const u32 digests_offset)
{
  /**
   * modifier
   */

  const u32 gid = get_global_id (0);
  const u32 lid = get_local_id (0);

  /**
   * salt
   */

  u32 salt_buf[8];

  salt_buf[0] = salt_bufs[salt_pos].salt_buf[0];
  salt_buf[1] = salt_bufs[salt_pos].salt_buf[1];
  salt_buf[2] = salt_bufs[salt_pos].salt_buf[2];
  salt_buf[3] = salt_bufs[salt_pos].salt_buf[3];
  salt_buf[4] = salt_bufs[salt_pos].salt_buf[4];
  salt_buf[5] = salt_bufs[salt_pos].salt_buf[5];
  salt_buf[6] = salt_bufs[salt_pos].salt_buf[6];
  salt_buf[7] = salt_bufs[salt_pos].salt_buf[7];

  const u32 salt_len = salt_bufs[salt_pos].salt_len;

  u32 s0[4];
  u32 s1[4];
  u32 s2[4];
  u32 s3[4];

  s0[0] = swap32 (salt_buf[0]);
  s0[1] = swap32 (salt_buf[1]);
  s0[2] = swap32 (salt_buf[2]);
  s0[3] = swap32 (salt_buf[3]);
  s1[0] = swap32 (salt_buf[4]);
  s1[1] = swap32 (salt_buf[5]);
  s1[2] = swap32 (salt_buf[6]);
  s1[3] = swap32 (salt_buf[7]);
  s2[0] = 0;
  s2[1] = 0;
  s2[2] = 0;
  s2[3] = 0;
  s3[0] = 0;
  s3[1] = 0;
  s3[2] = 0;
  s3[3] = 0;

  switch_buffer_by_offset_be_S (s0, s1, s2, s3, pw_len);

  const u32x pw_salt_len = pw_len + salt_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]
  };

  /**
   * loop
   */

  u32 w0l = w0[0];

  for (u32 il_pos = 0; il_pos < il_cnt; il_pos += VECT_SIZE)
  {
    const u32x w0r = ix_create_bft (bfs_buf, il_pos);

    const u32x w0lr = w0l | w0r;

    w0[0] = w0lr;

    /**
     * SAP
     */

    u32 final[32];

    final[ 0] = w0[0] | s0[0];
    final[ 1] = w0[1] | s0[1];
    final[ 2] = w0[2] | s0[2];
    final[ 3] = w0[3] | s0[3];
    final[ 4] = w1[0] | s1[0];
    final[ 5] = w1[1] | s1[1];
    final[ 6] = w1[2] | s1[2];
    final[ 7] = w1[3] | s1[3];
    final[ 8] = w2[0] | s2[0];
    final[ 9] = w2[1] | s2[1];
    final[10] = w2[2] | s2[2];
    final[11] = w2[3] | s2[3];
    final[12] = w3[0] | s3[0];
    final[13] = w3[1] | s3[1];
    final[14] = 0;
    final[15] = pw_salt_len * 8;

    u32 digest[5];

    digest[0] = SHA1M_A;
    digest[1] = SHA1M_B;
    digest[2] = SHA1M_C;
    digest[3] = SHA1M_D;
    digest[4] = SHA1M_E;

    sha1_transform (&final[0], &final[4], &final[8], &final[12], digest);

    // prepare magic array range

    u32 lengthMagicArray = 0x20;
    u32 offsetMagicArray = 0;

    lengthMagicArray += ((digest[0] >> 24) & 0xff) % 6;
    lengthMagicArray += ((digest[0] >> 16) & 0xff) % 6;
    lengthMagicArray += ((digest[0] >>  8) & 0xff) % 6;
    lengthMagicArray += ((digest[0] >>  0) & 0xff) % 6;
    lengthMagicArray += ((digest[1] >> 24) & 0xff) % 6;
    lengthMagicArray += ((digest[1] >> 16) & 0xff) % 6;
    lengthMagicArray += ((digest[1] >>  8) & 0xff) % 6;
    lengthMagicArray += ((digest[1] >>  0) & 0xff) % 6;
    lengthMagicArray += ((digest[2] >> 24) & 0xff) % 6;
    lengthMagicArray += ((digest[2] >> 16) & 0xff) % 6;
    offsetMagicArray += ((digest[2] >>  8) & 0xff) % 8;
    offsetMagicArray += ((digest[2] >>  0) & 0xff) % 8;
    offsetMagicArray += ((digest[3] >> 24) & 0xff) % 8;
    offsetMagicArray += ((digest[3] >> 16) & 0xff) % 8;
    offsetMagicArray += ((digest[3] >>  8) & 0xff) % 8;
    offsetMagicArray += ((digest[3] >>  0) & 0xff) % 8;
    offsetMagicArray += ((digest[4] >> 24) & 0xff) % 8;
    offsetMagicArray += ((digest[4] >> 16) & 0xff) % 8;
    offsetMagicArray += ((digest[4] >>  8) & 0xff) % 8;
    offsetMagicArray += ((digest[4] >>  0) & 0xff) % 8;

    // final

    digest[0] = SHA1M_A;
    digest[1] = SHA1M_B;
    digest[2] = SHA1M_C;
    digest[3] = SHA1M_D;
    digest[4] = SHA1M_E;

    #ifdef _unroll
    #pragma unroll
    #endif
    for (int i = 0; i < 32; i++) final[i] = 0;

    final[0] = swap32 (w0[0]);
    final[1] = swap32 (w0[1]);
    final[2] = swap32 (w0[2]);
    final[3] = swap32 (w0[3]);
    final[4] = swap32 (w1[0]);
    final[5] = swap32 (w1[1]);
    final[6] = swap32 (w1[2]);
    final[7] = swap32 (w1[3]);

    u32 final_len = pw_len;

    int i;

    // append MagicArray

    for (i = 0; i < lengthMagicArray - 4; i += 4)
    {
      const u32 tmp = GETSHIFTEDINT (theMagicArray, offsetMagicArray + i);

      SETSHIFTEDINT (final, final_len + i, tmp);
    }

    const u32 mask = 0xffffffff >> (((i - lengthMagicArray) & 3) * 8);

    const u32 tmp = GETSHIFTEDINT (theMagicArray, offsetMagicArray + i) & mask;

    SETSHIFTEDINT (final, final_len + i, tmp);

    final_len += lengthMagicArray;

    // append Salt

    for (i = 0; i < salt_len + 1; i += 4) // +1 for the 0x80
    {
      const u32 tmp = salt_buf[i / 4]; // attention, int[] not char[]

      SETSHIFTEDINT (final, final_len + i, tmp);
    }

    final_len += salt_len;

    // calculate

    int left;
    int off;

    for (left = final_len, off = 0; left >= 56; left -= 64, off += 16)
    {
      swap_buffer (&final[off]);

      sha1_transform (&final[off + 0], &final[off + 4], &final[off + 8], &final[off + 12], digest);
    }

    swap_buffer (&final[off]);

    final[off + 14] = 0;
    final[off + 15] = final_len * 8;

    sha1_transform (&final[off + 0], &final[off + 4], &final[off + 8], &final[off + 12], digest);

    COMPARE_S_SIMD (digest[3], digest[4], digest[2], digest[1]);
  }
}

__kernel void m07800_m04 (__global pw_t *pws, __global kernel_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 void *esalt_bufs, __global u32 *d_return_buf, __global u32 *d_scryptV0_buf, __global u32 *d_scryptV1_buf, __global u32 *d_scryptV2_buf, __global u32 *d_scryptV3_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 il_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
{
  /**
   * base
   */

  const u32 gid = get_global_id (0);

  if (gid >= gid_max) return;

  /**
   * modifier
   */

  //const u32 lid = get_local_id (0);

  u32 w0[4];

  w0[0] = pws[gid].i[ 0];
  w0[1] = pws[gid].i[ 1];
  w0[2] = pws[gid].i[ 2];
  w0[3] = pws[gid].i[ 3];

  u32 w1[4];

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

  u32 w2[4];

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

  u32 w3[4];

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

  const u32 pw_len = pws[gid].pw_len;

  /**
   * main
   */

  m07800m (w0, w1, w2, w3, pw_len, pws, rules_buf, combs_buf, bfs_buf, tmps, hooks, bitmaps_buf_s1_a, bitmaps_buf_s1_b, bitmaps_buf_s1_c, bitmaps_buf_s1_d, bitmaps_buf_s2_a, bitmaps_buf_s2_b, bitmaps_buf_s2_c, bitmaps_buf_s2_d, plains_buf, digests_buf, hashes_shown, salt_bufs, esalt_bufs, d_return_buf, d_scryptV0_buf, d_scryptV1_buf, d_scryptV2_buf, d_scryptV3_buf, bitmap_mask, bitmap_shift1, bitmap_shift2, salt_pos, loop_pos, loop_cnt, il_cnt, digests_cnt, digests_offset);
}

__kernel void m07800_m08 (__global pw_t *pws, __global kernel_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 void *esalt_bufs, __global u32 *d_return_buf, __global u32 *d_scryptV0_buf, __global u32 *d_scryptV1_buf, __global u32 *d_scryptV2_buf, __global u32 *d_scryptV3_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 il_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
{
  /**
   * base
   */

  const u32 gid = get_global_id (0);

  if (gid >= gid_max) return;

  /**
   * modifier
   */

  //const u32 lid = get_local_id (0);

  u32 w0[4];

  w0[0] = pws[gid].i[ 0];
  w0[1] = pws[gid].i[ 1];
  w0[2] = pws[gid].i[ 2];
  w0[3] = pws[gid].i[ 3];

  u32 w1[4];

  w1[0] = pws[gid].i[ 4];
  w1[1] = pws[gid].i[ 5];
  w1[2] = pws[gid].i[ 6];
  w1[3] = pws[gid].i[ 7];

  u32 w2[4];

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

  u32 w3[4];

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

  const u32 pw_len = pws[gid].pw_len;

  /**
   * main
   */

  m07800m (w0, w1, w2, w3, pw_len, pws, rules_buf, combs_buf, bfs_buf, tmps, hooks, bitmaps_buf_s1_a, bitmaps_buf_s1_b, bitmaps_buf_s1_c, bitmaps_buf_s1_d, bitmaps_buf_s2_a, bitmaps_buf_s2_b, bitmaps_buf_s2_c, bitmaps_buf_s2_d, plains_buf, digests_buf, hashes_shown, salt_bufs, esalt_bufs, d_return_buf, d_scryptV0_buf, d_scryptV1_buf, d_scryptV2_buf, d_scryptV3_buf, bitmap_mask, bitmap_shift1, bitmap_shift2, salt_pos, loop_pos, loop_cnt, il_cnt, digests_cnt, digests_offset);
}

__kernel void m07800_m16 (__global pw_t *pws, __global kernel_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 void *esalt_bufs, __global u32 *d_return_buf, __global u32 *d_scryptV0_buf, __global u32 *d_scryptV1_buf, __global u32 *d_scryptV2_buf, __global u32 *d_scryptV3_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 il_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
{
}

__kernel void m07800_s04 (__global pw_t *pws, __global kernel_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 void *esalt_bufs, __global u32 *d_return_buf, __global u32 *d_scryptV0_buf, __global u32 *d_scryptV1_buf, __global u32 *d_scryptV2_buf, __global u32 *d_scryptV3_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 il_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
{
  /**
   * base
   */

  const u32 gid = get_global_id (0);

  if (gid >= gid_max) return;

  /**
   * modifier
   */

  //const u32 lid = get_local_id (0);

  u32 w0[4];

  w0[0] = pws[gid].i[ 0];
  w0[1] = pws[gid].i[ 1];
  w0[2] = pws[gid].i[ 2];
  w0[3] = pws[gid].i[ 3];

  u32 w1[4];

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

  u32 w2[4];

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

  u32 w3[4];

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

  const u32 pw_len = pws[gid].pw_len;

  /**
   * main
   */

  m07800s (w0, w1, w2, w3, pw_len, pws, rules_buf, combs_buf, bfs_buf, tmps, hooks, bitmaps_buf_s1_a, bitmaps_buf_s1_b, bitmaps_buf_s1_c, bitmaps_buf_s1_d, bitmaps_buf_s2_a, bitmaps_buf_s2_b, bitmaps_buf_s2_c, bitmaps_buf_s2_d, plains_buf, digests_buf, hashes_shown, salt_bufs, esalt_bufs, d_return_buf, d_scryptV0_buf, d_scryptV1_buf, d_scryptV2_buf, d_scryptV3_buf, bitmap_mask, bitmap_shift1, bitmap_shift2, salt_pos, loop_pos, loop_cnt, il_cnt, digests_cnt, digests_offset);
}

__kernel void m07800_s08 (__global pw_t *pws, __global kernel_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 void *esalt_bufs, __global u32 *d_return_buf, __global u32 *d_scryptV0_buf, __global u32 *d_scryptV1_buf, __global u32 *d_scryptV2_buf, __global u32 *d_scryptV3_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 il_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
{
  /**
   * base
   */

  const u32 gid = get_global_id (0);

  if (gid >= gid_max) return;

  /**
   * modifier
   */

  //const u32 lid = get_local_id (0);

  u32 w0[4];

  w0[0] = pws[gid].i[ 0];
  w0[1] = pws[gid].i[ 1];
  w0[2] = pws[gid].i[ 2];
  w0[3] = pws[gid].i[ 3];

  u32 w1[4];

  w1[0] = pws[gid].i[ 4];
  w1[1] = pws[gid].i[ 5];
  w1[2] = pws[gid].i[ 6];
  w1[3] = pws[gid].i[ 7];

  u32 w2[4];

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

  u32 w3[4];

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

  const u32 pw_len = pws[gid].pw_len;

  /**
   * main
   */

  m07800s (w0, w1, w2, w3, pw_len, pws, rules_buf, combs_buf, bfs_buf, tmps, hooks, bitmaps_buf_s1_a, bitmaps_buf_s1_b, bitmaps_buf_s1_c, bitmaps_buf_s1_d, bitmaps_buf_s2_a, bitmaps_buf_s2_b, bitmaps_buf_s2_c, bitmaps_buf_s2_d, plains_buf, digests_buf, hashes_shown, salt_bufs, esalt_bufs, d_return_buf, d_scryptV0_buf, d_scryptV1_buf, d_scryptV2_buf, d_scryptV3_buf, bitmap_mask, bitmap_shift1, bitmap_shift2, salt_pos, loop_pos, loop_cnt, il_cnt, digests_cnt, digests_offset);
}

__kernel void m07800_s16 (__global pw_t *pws, __global kernel_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 void *esalt_bufs, __global u32 *d_return_buf, __global u32 *d_scryptV0_buf, __global u32 *d_scryptV1_buf, __global u32 *d_scryptV2_buf, __global u32 *d_scryptV3_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 il_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
{
}