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

#define _SHA256_

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

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

#include "include/kernel_functions.c"
#include "OpenCL/types_ocl.c"
#include "OpenCL/common.c"

#define COMPARE_S "OpenCL/check_single_comp4.c"
#define COMPARE_M "OpenCL/check_multi_comp4.c"

__constant u32 k_sha256[64] =
{
  SHA256C00, SHA256C01, SHA256C02, SHA256C03,
  SHA256C04, SHA256C05, SHA256C06, SHA256C07,
  SHA256C08, SHA256C09, SHA256C0a, SHA256C0b,
  SHA256C0c, SHA256C0d, SHA256C0e, SHA256C0f,
  SHA256C10, SHA256C11, SHA256C12, SHA256C13,
  SHA256C14, SHA256C15, SHA256C16, SHA256C17,
  SHA256C18, SHA256C19, SHA256C1a, SHA256C1b,
  SHA256C1c, SHA256C1d, SHA256C1e, SHA256C1f,
  SHA256C20, SHA256C21, SHA256C22, SHA256C23,
  SHA256C24, SHA256C25, SHA256C26, SHA256C27,
  SHA256C28, SHA256C29, SHA256C2a, SHA256C2b,
  SHA256C2c, SHA256C2d, SHA256C2e, SHA256C2f,
  SHA256C30, SHA256C31, SHA256C32, SHA256C33,
  SHA256C34, SHA256C35, SHA256C36, SHA256C37,
  SHA256C38, SHA256C39, SHA256C3a, SHA256C3b,
  SHA256C3c, SHA256C3d, SHA256C3e, SHA256C3f,
};

static void sha256_transform (const u32 w0[4], const u32 w1[4], const u32 w2[4], const u32 w3[4], u32 digest[8])
{
  u32 a = digest[0];
  u32 b = digest[1];
  u32 c = digest[2];
  u32 d = digest[3];
  u32 e = digest[4];
  u32 f = digest[5];
  u32 g = digest[6];
  u32 h = digest[7];

  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];

  #define ROUND_EXPAND()                            \
  {                                                 \
    w0_t = SHA256_EXPAND (we_t, w9_t, w1_t, w0_t);  \
    w1_t = SHA256_EXPAND (wf_t, wa_t, w2_t, w1_t);  \
    w2_t = SHA256_EXPAND (w0_t, wb_t, w3_t, w2_t);  \
    w3_t = SHA256_EXPAND (w1_t, wc_t, w4_t, w3_t);  \
    w4_t = SHA256_EXPAND (w2_t, wd_t, w5_t, w4_t);  \
    w5_t = SHA256_EXPAND (w3_t, we_t, w6_t, w5_t);  \
    w6_t = SHA256_EXPAND (w4_t, wf_t, w7_t, w6_t);  \
    w7_t = SHA256_EXPAND (w5_t, w0_t, w8_t, w7_t);  \
    w8_t = SHA256_EXPAND (w6_t, w1_t, w9_t, w8_t);  \
    w9_t = SHA256_EXPAND (w7_t, w2_t, wa_t, w9_t);  \
    wa_t = SHA256_EXPAND (w8_t, w3_t, wb_t, wa_t);  \
    wb_t = SHA256_EXPAND (w9_t, w4_t, wc_t, wb_t);  \
    wc_t = SHA256_EXPAND (wa_t, w5_t, wd_t, wc_t);  \
    wd_t = SHA256_EXPAND (wb_t, w6_t, we_t, wd_t);  \
    we_t = SHA256_EXPAND (wc_t, w7_t, wf_t, we_t);  \
    wf_t = SHA256_EXPAND (wd_t, w8_t, w0_t, wf_t);  \
  }

  #define ROUND_STEP(i)                                                                   \
  {                                                                                       \
    SHA256_STEP (SHA256_F0o, SHA256_F1o, a, b, c, d, e, f, g, h, w0_t, k_sha256[i +  0]); \
    SHA256_STEP (SHA256_F0o, SHA256_F1o, h, a, b, c, d, e, f, g, w1_t, k_sha256[i +  1]); \
    SHA256_STEP (SHA256_F0o, SHA256_F1o, g, h, a, b, c, d, e, f, w2_t, k_sha256[i +  2]); \
    SHA256_STEP (SHA256_F0o, SHA256_F1o, f, g, h, a, b, c, d, e, w3_t, k_sha256[i +  3]); \
    SHA256_STEP (SHA256_F0o, SHA256_F1o, e, f, g, h, a, b, c, d, w4_t, k_sha256[i +  4]); \
    SHA256_STEP (SHA256_F0o, SHA256_F1o, d, e, f, g, h, a, b, c, w5_t, k_sha256[i +  5]); \
    SHA256_STEP (SHA256_F0o, SHA256_F1o, c, d, e, f, g, h, a, b, w6_t, k_sha256[i +  6]); \
    SHA256_STEP (SHA256_F0o, SHA256_F1o, b, c, d, e, f, g, h, a, w7_t, k_sha256[i +  7]); \
    SHA256_STEP (SHA256_F0o, SHA256_F1o, a, b, c, d, e, f, g, h, w8_t, k_sha256[i +  8]); \
    SHA256_STEP (SHA256_F0o, SHA256_F1o, h, a, b, c, d, e, f, g, w9_t, k_sha256[i +  9]); \
    SHA256_STEP (SHA256_F0o, SHA256_F1o, g, h, a, b, c, d, e, f, wa_t, k_sha256[i + 10]); \
    SHA256_STEP (SHA256_F0o, SHA256_F1o, f, g, h, a, b, c, d, e, wb_t, k_sha256[i + 11]); \
    SHA256_STEP (SHA256_F0o, SHA256_F1o, e, f, g, h, a, b, c, d, wc_t, k_sha256[i + 12]); \
    SHA256_STEP (SHA256_F0o, SHA256_F1o, d, e, f, g, h, a, b, c, wd_t, k_sha256[i + 13]); \
    SHA256_STEP (SHA256_F0o, SHA256_F1o, c, d, e, f, g, h, a, b, we_t, k_sha256[i + 14]); \
    SHA256_STEP (SHA256_F0o, SHA256_F1o, b, c, d, e, f, g, h, a, wf_t, k_sha256[i + 15]); \
  }

  ROUND_STEP (0);

  #pragma unroll
  for (int i = 16; i < 64; i += 16)
  {
    ROUND_EXPAND (); ROUND_STEP (i);
  }

  digest[0] += a;
  digest[1] += b;
  digest[2] += c;
  digest[3] += d;
  digest[4] += e;
  digest[5] += f;
  digest[6] += g;
  digest[7] += h;
}

__kernel void m05200_init (__global pw_t *pws, __global kernel_rule_t *rules_buf, __global comb_t *combs_buf, __global bf_t *bfs_buf, __global pwsafe3_tmp_t *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_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)
{
  /**
   * base
   */

  const u32 gid = get_global_id (0);

  if (gid >= gid_max) return;

  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] = 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;

  /**
   * salt
   */

  u32 salt_buf0[4];

  salt_buf0[0] = salt_bufs[salt_pos].salt_buf[0];
  salt_buf0[1] = salt_bufs[salt_pos].salt_buf[1];
  salt_buf0[2] = salt_bufs[salt_pos].salt_buf[2];
  salt_buf0[3] = salt_bufs[salt_pos].salt_buf[3];

  u32 salt_buf1[4];

  salt_buf1[0] = salt_bufs[salt_pos].salt_buf[4];
  salt_buf1[1] = salt_bufs[salt_pos].salt_buf[5];
  salt_buf1[2] = salt_bufs[salt_pos].salt_buf[6];
  salt_buf1[3] = salt_bufs[salt_pos].salt_buf[7];

  u32 salt_buf2[4];

  salt_buf2[0] = 0;
  salt_buf2[1] = 0;
  salt_buf2[2] = 0;
  salt_buf2[3] = 0;

  u32 salt_buf3[4];

  salt_buf3[0] = 0;
  salt_buf3[1] = 0;
  salt_buf3[2] = 0;
  salt_buf3[3] = 0;

  u32 salt_len = salt_bufs[salt_pos].salt_len;

  switch_buffer_by_offset (salt_buf0, salt_buf1, salt_buf2, salt_buf3, pw_len);

  w0[0] |= salt_buf0[0];
  w0[1] |= salt_buf0[1];
  w0[2] |= salt_buf0[2];
  w0[3] |= salt_buf0[3];

  w1[0] |= salt_buf1[0];
  w1[1] |= salt_buf1[1];
  w1[2] |= salt_buf1[2];
  w1[3] |= salt_buf1[3];

  w2[0] |= salt_buf2[0];
  w2[1] |= salt_buf2[1];
  w2[2] |= salt_buf2[2];
  w2[3] |= salt_buf2[3];

  w3[0] |= salt_buf3[0];
  w3[1] |= salt_buf3[1];
  w3[2] |= salt_buf3[2];
  w3[3] |= salt_buf3[3];

  const u32 block_len = pw_len + salt_len;

  append_0x80_4x4 (w0, w1, w2, w3, block_len);

  /**
   * init
   */

  w0[0] = swap32 (w0[0]);
  w0[1] = swap32 (w0[1]);
  w0[2] = swap32 (w0[2]);
  w0[3] = swap32 (w0[3]);

  w1[0] = swap32 (w1[0]);
  w1[1] = swap32 (w1[1]);
  w1[2] = swap32 (w1[2]);
  w1[3] = swap32 (w1[3]);

  w2[0] = swap32 (w2[0]);
  w2[1] = swap32 (w2[1]);
  w2[2] = swap32 (w2[2]);
  w2[3] = swap32 (w2[3]);

  w3[0] = swap32 (w3[0]);
  w3[1] = swap32 (w3[1]);
  w3[2] = 0;
  w3[3] = block_len * 8;

  /**
   * main
   */

  u32 digest[8];

  digest[0] = SHA256M_A;
  digest[1] = SHA256M_B;
  digest[2] = SHA256M_C;
  digest[3] = SHA256M_D;
  digest[4] = SHA256M_E;
  digest[5] = SHA256M_F;
  digest[6] = SHA256M_G;
  digest[7] = SHA256M_H;

  sha256_transform (w0, w1, w2, w3, digest);

  /**
   * store
   */

  tmps[gid].digest_buf[0] = digest[0];
  tmps[gid].digest_buf[1] = digest[1];
  tmps[gid].digest_buf[2] = digest[2];
  tmps[gid].digest_buf[3] = digest[3];
  tmps[gid].digest_buf[4] = digest[4];
  tmps[gid].digest_buf[5] = digest[5];
  tmps[gid].digest_buf[6] = digest[6];
  tmps[gid].digest_buf[7] = digest[7];
}

__kernel void m05200_loop (__global pw_t *pws, __global kernel_rule_t *rules_buf, __global comb_t *combs_buf, __global bf_t *bfs_buf, __global pwsafe3_tmp_t *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_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)
{
  const u32 gid = get_global_id (0);

  if (gid >= gid_max) return;

  u32 digest[8];

  digest[0] = tmps[gid].digest_buf[0];
  digest[1] = tmps[gid].digest_buf[1];
  digest[2] = tmps[gid].digest_buf[2];
  digest[3] = tmps[gid].digest_buf[3];
  digest[4] = tmps[gid].digest_buf[4];
  digest[5] = tmps[gid].digest_buf[5];
  digest[6] = tmps[gid].digest_buf[6];
  digest[7] = tmps[gid].digest_buf[7];

  /**
   * init
   */

  for (u32 i = 0; i < loop_cnt; i++)
  {
    u32 w0[4];

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

    u32 w1[4];

    w1[0] = digest[4];
    w1[1] = digest[5];
    w1[2] = digest[6];
    w1[3] = digest[7];

    u32 w2[4];

    w2[0] = 0x80000000;
    w2[1] = 0;
    w2[2] = 0;
    w2[3] = 0;

    u32 w3[4];

    w3[0] = 0;
    w3[1] = 0;
    w3[2] = 0;
    w3[3] = 32 * 8;

    digest[0] = SHA256M_A;
    digest[1] = SHA256M_B;
    digest[2] = SHA256M_C;
    digest[3] = SHA256M_D;
    digest[4] = SHA256M_E;
    digest[5] = SHA256M_F;
    digest[6] = SHA256M_G;
    digest[7] = SHA256M_H;

    sha256_transform (w0, w1, w2, w3, digest);
  }

  tmps[gid].digest_buf[0] = digest[0];
  tmps[gid].digest_buf[1] = digest[1];
  tmps[gid].digest_buf[2] = digest[2];
  tmps[gid].digest_buf[3] = digest[3];
  tmps[gid].digest_buf[4] = digest[4];
  tmps[gid].digest_buf[5] = digest[5];
  tmps[gid].digest_buf[6] = digest[6];
  tmps[gid].digest_buf[7] = digest[7];
}

__kernel void m05200_comp (__global pw_t *pws, __global kernel_rule_t *rules_buf, __global comb_t *combs_buf, __global bf_t *bfs_buf, __global pwsafe3_tmp_t *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_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 gid = get_global_id (0);

  if (gid >= gid_max) return;

  const u32 lid = get_local_id (0);

  /**
   * digest
   */

  const u32 r0 = tmps[gid].digest_buf[DGST_R0];
  const u32 r1 = tmps[gid].digest_buf[DGST_R1];
  const u32 r2 = tmps[gid].digest_buf[DGST_R2];
  const u32 r3 = tmps[gid].digest_buf[DGST_R3];

  #define il_pos 0

  #include COMPARE_M
}