#define _KRB5PA_
-#include "include/constants.h"
-#include "include/kernel_vendor.h"
+//shared mem too small
+//#define NEW_SIMD_CODE
-#define DGST_R0 3
-#define DGST_R1 7
-#define DGST_R2 2
-#define DGST_R3 6
-
-#include "include/kernel_functions.c"
-#include "types_ocl.c"
-#include "common.c"
+#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"
typedef struct
{
u8 S[256];
- u8 i;
- u8 j;
+ u32 wtf_its_faster;
} RC4_KEY;
-static void swap (__local RC4_KEY *rc4_key, const u8 i, const u8 j)
+void swap (__local RC4_KEY *rc4_key, const u8 i, const u8 j)
{
u8 tmp;
rc4_key->S[j] = tmp;
}
-static void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
+void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
{
- u32 i;
+ u32 v = 0x03020100;
+ u32 a = 0x04040404;
- #pragma unroll 256
- for (i = 0; i < 256; i += 1) rc4_key->S[i] = i;
+ __local u32 *ptr = (__local u32 *) rc4_key->S;
- u8 j = 0;
+ #ifdef _unroll
+ #pragma unroll
+ #endif
+ for (u32 i = 0; i < 64; i++)
+ {
+ *ptr++ = v; v += a;
+ }
+
+ u32 j = 0;
- #pragma unroll 16
- for (i = 0; i < 256; i += 16)
+ for (u32 i = 0; i < 16; i++)
{
- u32 idx = i;
+ u32 idx = i * 16;
u32 v;
v = data[0];
- j += rc4_key->S[idx] + (v >> 0); swap (rc4_key, idx, j); idx++;
- j += rc4_key->S[idx] + (v >> 8); swap (rc4_key, idx, j); idx++;
- j += rc4_key->S[idx] + (v >> 16); swap (rc4_key, idx, j); idx++;
- j += rc4_key->S[idx] + (v >> 24); swap (rc4_key, idx, j); idx++;
+ j += rc4_key->S[idx] + (v >> 0); swap (rc4_key, idx, j); idx++;
+ j += rc4_key->S[idx] + (v >> 8); swap (rc4_key, idx, j); idx++;
+ j += rc4_key->S[idx] + (v >> 16); swap (rc4_key, idx, j); idx++;
+ j += rc4_key->S[idx] + (v >> 24); swap (rc4_key, idx, j); idx++;
v = data[1];
- j += rc4_key->S[idx] + (v >> 0); swap (rc4_key, idx, j); idx++;
- j += rc4_key->S[idx] + (v >> 8); swap (rc4_key, idx, j); idx++;
- j += rc4_key->S[idx] + (v >> 16); swap (rc4_key, idx, j); idx++;
- j += rc4_key->S[idx] + (v >> 24); swap (rc4_key, idx, j); idx++;
+ j += rc4_key->S[idx] + (v >> 0); swap (rc4_key, idx, j); idx++;
+ j += rc4_key->S[idx] + (v >> 8); swap (rc4_key, idx, j); idx++;
+ j += rc4_key->S[idx] + (v >> 16); swap (rc4_key, idx, j); idx++;
+ j += rc4_key->S[idx] + (v >> 24); swap (rc4_key, idx, j); idx++;
v = data[2];
- j += rc4_key->S[idx] + (v >> 0); swap (rc4_key, idx, j); idx++;
- j += rc4_key->S[idx] + (v >> 8); swap (rc4_key, idx, j); idx++;
- j += rc4_key->S[idx] + (v >> 16); swap (rc4_key, idx, j); idx++;
- j += rc4_key->S[idx] + (v >> 24); swap (rc4_key, idx, j); idx++;
+ j += rc4_key->S[idx] + (v >> 0); swap (rc4_key, idx, j); idx++;
+ j += rc4_key->S[idx] + (v >> 8); swap (rc4_key, idx, j); idx++;
+ j += rc4_key->S[idx] + (v >> 16); swap (rc4_key, idx, j); idx++;
+ j += rc4_key->S[idx] + (v >> 24); swap (rc4_key, idx, j); idx++;
v = data[3];
- j += rc4_key->S[idx] + (v >> 0); swap (rc4_key, idx, j); idx++;
- j += rc4_key->S[idx] + (v >> 8); swap (rc4_key, idx, j); idx++;
- j += rc4_key->S[idx] + (v >> 16); swap (rc4_key, idx, j); idx++;
- j += rc4_key->S[idx] + (v >> 24); swap (rc4_key, idx, j); idx++;
+ j += rc4_key->S[idx] + (v >> 0); swap (rc4_key, idx, j); idx++;
+ j += rc4_key->S[idx] + (v >> 8); swap (rc4_key, idx, j); idx++;
+ j += rc4_key->S[idx] + (v >> 16); swap (rc4_key, idx, j); idx++;
+ j += rc4_key->S[idx] + (v >> 24); swap (rc4_key, idx, j); idx++;
}
-
- rc4_key->i = 0;
- rc4_key->j = 0;
}
-static u32 rc4_next_4 (__local RC4_KEY *rc4_key, const u32 ct)
+u8 rc4_next_16 (__local RC4_KEY *rc4_key, u8 i, u8 j, const u32 in[4], u32 out[4])
{
- u8 idx;
-
- u32 xor4 = 0;
+ #ifdef _unroll
+ #pragma unroll
+ #endif
+ for (u32 k = 0; k < 4; k++)
+ {
+ u32 xor4 = 0;
- u8 i = rc4_key->i;
- u8 j = rc4_key->j;
+ u8 idx;
- i += 1;
- j += rc4_key->S[i];
+ i += 1;
+ j += rc4_key->S[i];
- swap (rc4_key, i, j);
+ swap (rc4_key, i, j);
- idx = rc4_key->S[i] + rc4_key->S[j];
+ idx = rc4_key->S[i] + rc4_key->S[j];
- xor4 |= rc4_key->S[idx] << 0;
+ xor4 |= rc4_key->S[idx] << 0;
- i += 1;
- j += rc4_key->S[i];
+ i += 1;
+ j += rc4_key->S[i];
- swap (rc4_key, i, j);
+ swap (rc4_key, i, j);
- idx = rc4_key->S[i] + rc4_key->S[j];
+ idx = rc4_key->S[i] + rc4_key->S[j];
- xor4 |= rc4_key->S[idx] << 8;
+ xor4 |= rc4_key->S[idx] << 8;
- i += 1;
- j += rc4_key->S[i];
+ i += 1;
+ j += rc4_key->S[i];
- swap (rc4_key, i, j);
+ swap (rc4_key, i, j);
- idx = rc4_key->S[i] + rc4_key->S[j];
+ idx = rc4_key->S[i] + rc4_key->S[j];
- xor4 |= rc4_key->S[idx] << 16;
+ xor4 |= rc4_key->S[idx] << 16;
- i += 1;
- j += rc4_key->S[i];
+ i += 1;
+ j += rc4_key->S[i];
- swap (rc4_key, i, j);
+ swap (rc4_key, i, j);
- idx = rc4_key->S[i] + rc4_key->S[j];
+ idx = rc4_key->S[i] + rc4_key->S[j];
- xor4 |= rc4_key->S[idx] << 24;
+ xor4 |= rc4_key->S[idx] << 24;
- rc4_key->i = i;
- rc4_key->j = j;
+ out[k] = in[k] ^ xor4;
+ }
- return ct ^ xor4;
+ return j;
}
-static int decrypt_and_check (__local RC4_KEY *rc4_key, u32 data[4], u32 timestamp_ct[8])
+int decrypt_and_check (__local RC4_KEY *rc4_key, u32 data[4], u32 timestamp_ct[8])
{
- u32 pt;
-
rc4_init_16 (rc4_key, data);
- pt = rc4_next_4 (rc4_key, timestamp_ct[0]);
- pt = rc4_next_4 (rc4_key, timestamp_ct[1]);
- pt = rc4_next_4 (rc4_key, timestamp_ct[2]);
- pt = rc4_next_4 (rc4_key, timestamp_ct[3]);
-
- if ((pt & 0xffff0000) != 0x30320000) return 0;
-
- pt = rc4_next_4 (rc4_key, timestamp_ct[4]);
+ u32 out[4];
- if (((pt & 0xff) < '0') || ((pt & 0xff) > '9')) return 0; pt >>= 8;
- if (((pt & 0xff) < '0') || ((pt & 0xff) > '9')) return 0; pt >>= 8;
- if (((pt & 0xff) < '0') || ((pt & 0xff) > '9')) return 0; pt >>= 8;
- if (((pt & 0xff) < '0') || ((pt & 0xff) > '9')) return 0;
+ u8 j = 0;
- pt = rc4_next_4 (rc4_key, timestamp_ct[5]);
+ j = rc4_next_16 (rc4_key, 0, j, timestamp_ct + 0, out);
- if (((pt & 0xff) < '0') || ((pt & 0xff) > '9')) return 0; pt >>= 8;
- if (((pt & 0xff) < '0') || ((pt & 0xff) > '9')) return 0; pt >>= 8;
- if (((pt & 0xff) < '0') || ((pt & 0xff) > '9')) return 0; pt >>= 8;
- if (((pt & 0xff) < '0') || ((pt & 0xff) > '9')) return 0;
+ if ((out[3] & 0xffff0000) != 0x30320000) return 0;
- pt = rc4_next_4 (rc4_key, timestamp_ct[6]);
+ j = rc4_next_16 (rc4_key, 16, j, timestamp_ct + 4, out);
- if (((pt & 0xff) < '0') || ((pt & 0xff) > '9')) return 0; pt >>= 8;
- if (((pt & 0xff) < '0') || ((pt & 0xff) > '9')) return 0; pt >>= 8;
- if (((pt & 0xff) < '0') || ((pt & 0xff) > '9')) return 0; pt >>= 8;
- if (((pt & 0xff) < '0') || ((pt & 0xff) > '9')) return 0;
+ if (((out[0] & 0xff) < '0') || ((out[0] & 0xff) > '9')) return 0; out[0] >>= 8;
+ if (((out[0] & 0xff) < '0') || ((out[0] & 0xff) > '9')) return 0; out[0] >>= 8;
+ if (((out[0] & 0xff) < '0') || ((out[0] & 0xff) > '9')) return 0; out[0] >>= 8;
+ if (((out[0] & 0xff) < '0') || ((out[0] & 0xff) > '9')) return 0;
+ if (((out[1] & 0xff) < '0') || ((out[1] & 0xff) > '9')) return 0; out[1] >>= 8;
+ if (((out[1] & 0xff) < '0') || ((out[1] & 0xff) > '9')) return 0; out[1] >>= 8;
+ if (((out[1] & 0xff) < '0') || ((out[1] & 0xff) > '9')) return 0; out[1] >>= 8;
+ if (((out[1] & 0xff) < '0') || ((out[1] & 0xff) > '9')) return 0;
+ if (((out[2] & 0xff) < '0') || ((out[2] & 0xff) > '9')) return 0; out[2] >>= 8;
+ if (((out[2] & 0xff) < '0') || ((out[2] & 0xff) > '9')) return 0; out[2] >>= 8;
+ if (((out[2] & 0xff) < '0') || ((out[2] & 0xff) > '9')) return 0; out[2] >>= 8;
+ if (((out[2] & 0xff) < '0') || ((out[2] & 0xff) > '9')) return 0;
return 1;
}
-static void md4_transform (const u32 w0[4], const u32 w1[4], const u32 w2[4], const u32 w3[4], u32 digest[4])
+void md4_transform (const u32 w0[4], const u32 w1[4], const u32 w2[4], const u32 w3[4], u32 digest[4])
{
u32 a = digest[0];
u32 b = digest[1];
digest[3] += d;
}
-static void md5_transform (const u32 w0[4], const u32 w1[4], const u32 w2[4], const u32 w3[4], u32 digest[4])
+void md5_transform (const u32 w0[4], const u32 w1[4], const u32 w2[4], const u32 w3[4], u32 digest[4])
{
u32 a = digest[0];
u32 b = digest[1];
digest[3] += d;
}
-static void hmac_md5_pad (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], u32 ipad[4], u32 opad[4])
+void hmac_md5_pad (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], u32 ipad[4], u32 opad[4])
{
w0[0] = w0[0] ^ 0x36363636;
w0[1] = w0[1] ^ 0x36363636;
md5_transform (w0, w1, w2, w3, opad);
}
-static void hmac_md5_run (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], u32 ipad[4], u32 opad[4], u32 digest[4])
+void hmac_md5_run (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], u32 ipad[4], u32 opad[4], u32 digest[4])
{
digest[0] = ipad[0];
digest[1] = ipad[1];
md5_transform (w0, w1, w2, w3, digest);
}
-static void kerb_prepare (const u32 w0[4], const u32 w1[4], const u32 pw_len, const u32 checksum[4], u32 digest[4])
+void kerb_prepare (const u32 w0[4], const u32 w1[4], const u32 pw_len, const u32 checksum[4], u32 digest[4])
{
/**
* pads
hmac_md5_run (w0_t, w1_t, w2_t, w3_t, ipad, opad, digest);
}
-__kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m07500_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 krb5pa_t *krb5pa_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 combs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
+__kernel void m07500_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 krb5pa_t *krb5pa_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 il_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
{
/**
* modifier
*/
- __local RC4_KEY rc4_keys[64];
-
- const u32 gid = get_global_id (0);
const u32 lid = get_local_id (0);
- if (gid >= gid_max) return;
-
/**
* base
*/
- u32 wordl0[4];
-
- wordl0[0] = pws[gid].i[ 0];
- wordl0[1] = pws[gid].i[ 1];
- wordl0[2] = pws[gid].i[ 2];
- wordl0[3] = pws[gid].i[ 3];
-
- u32 wordl1[4];
-
- wordl1[0] = pws[gid].i[ 4];
- wordl1[1] = pws[gid].i[ 5];
- wordl1[2] = pws[gid].i[ 6];
- wordl1[3] = pws[gid].i[ 7];
-
- u32 wordl2[4];
+ const u32 gid = get_global_id (0);
- wordl2[0] = 0;
- wordl2[1] = 0;
- wordl2[2] = 0;
- wordl2[3] = 0;
+ if (gid >= gid_max) return;
- u32 wordl3[4];
+ u32 pw_buf0[4];
+ u32 pw_buf1[4];
- wordl3[0] = 0;
- wordl3[1] = 0;
- wordl3[2] = 0;
- wordl3[3] = 0;
+ 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];
+ 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_l_len = pws[gid].pw_len;
- if (combs_mode == COMBINATOR_MODE_BASE_RIGHT)
- {
- switch_buffer_by_offset (wordl0, wordl1, wordl2, wordl3, combs_buf[0].pw_len);
- }
-
/**
* salt
*/
timestamp_ct[7] = krb5pa_bufs[salt_pos].timestamp[7];
/**
- * loop
+ * shared
*/
- for (u32 il_pos = 0; il_pos < combs_cnt; il_pos++)
- {
- const u32 pw_r_len = combs_buf[il_pos].pw_len;
-
- const u32 pw_len = pw_l_len + pw_r_len;
-
- u32 wordr0[4];
-
- wordr0[0] = combs_buf[il_pos].i[0];
- wordr0[1] = combs_buf[il_pos].i[1];
- wordr0[2] = combs_buf[il_pos].i[2];
- wordr0[3] = combs_buf[il_pos].i[3];
-
- u32 wordr1[4];
+ __local RC4_KEY rc4_keys[64];
- wordr1[0] = combs_buf[il_pos].i[4];
- wordr1[1] = combs_buf[il_pos].i[5];
- wordr1[2] = combs_buf[il_pos].i[6];
- wordr1[3] = combs_buf[il_pos].i[7];
+ /**
+ * loop
+ */
- u32 wordr2[4];
+ for (u32 il_pos = 0; il_pos < il_cnt; il_pos += VECT_SIZE)
+ {
+ const u32x pw_r_len = pwlenx_create_combt (combs_buf, il_pos);
- wordr2[0] = 0;
- wordr2[1] = 0;
- wordr2[2] = 0;
- wordr2[3] = 0;
+ const u32x pw_len = pw_l_len + pw_r_len;
- u32 wordr3[4];
+ /**
+ * concat password candidate
+ */
- wordr3[0] = 0;
- wordr3[1] = 0;
- wordr3[2] = 0;
- wordr3[3] = 0;
+ u32x wordl0[4] = { 0 };
+ u32x wordl1[4] = { 0 };
+ u32x wordl2[4] = { 0 };
+ u32x wordl3[4] = { 0 };
+
+ wordl0[0] = pw_buf0[0];
+ wordl0[1] = pw_buf0[1];
+ wordl0[2] = pw_buf0[2];
+ wordl0[3] = pw_buf0[3];
+ wordl1[0] = pw_buf1[0];
+ wordl1[1] = pw_buf1[1];
+ wordl1[2] = pw_buf1[2];
+ wordl1[3] = pw_buf1[3];
+
+ u32x wordr0[4] = { 0 };
+ u32x wordr1[4] = { 0 };
+ u32x wordr2[4] = { 0 };
+ u32x wordr3[4] = { 0 };
+
+ wordr0[0] = ix_create_combt (combs_buf, il_pos, 0);
+ wordr0[1] = ix_create_combt (combs_buf, il_pos, 1);
+ wordr0[2] = ix_create_combt (combs_buf, il_pos, 2);
+ wordr0[3] = ix_create_combt (combs_buf, il_pos, 3);
+ wordr1[0] = ix_create_combt (combs_buf, il_pos, 4);
+ wordr1[1] = ix_create_combt (combs_buf, il_pos, 5);
+ wordr1[2] = ix_create_combt (combs_buf, il_pos, 6);
+ wordr1[3] = ix_create_combt (combs_buf, il_pos, 7);
if (combs_mode == COMBINATOR_MODE_BASE_LEFT)
{
- switch_buffer_by_offset (wordr0, wordr1, wordr2, wordr3, pw_l_len);
+ switch_buffer_by_offset_le_VV (wordr0, wordr1, wordr2, wordr3, pw_l_len);
+ }
+ else
+ {
+ switch_buffer_by_offset_le_VV (wordl0, wordl1, wordl2, wordl3, pw_r_len);
}
- u32 w0[4];
+ u32x w0[4];
+ u32x w1[4];
w0[0] = wordl0[0] | wordr0[0];
w0[1] = wordl0[1] | wordr0[1];
w0[2] = wordl0[2] | wordr0[2];
w0[3] = wordl0[3] | wordr0[3];
-
- u32 w1[4];
-
w1[0] = wordl1[0] | wordr1[0];
w1[1] = wordl1[1] | wordr1[1];
w1[2] = wordl1[2] | wordr1[2];
w1[3] = wordl1[3] | wordr1[3];
- u32 w2[4];
-
- w2[0] = wordl2[0] | wordr2[0];
- w2[1] = wordl2[1] | wordr2[1];
- w2[2] = wordl2[2] | wordr2[2];
- w2[3] = wordl2[3] | wordr2[3];
-
- u32 w3[4];
-
- w3[0] = wordl3[0] | wordr3[0];
- w3[1] = wordl3[1] | wordr3[1];
- w3[2] = 0;
- w3[3] = 0;
-
/**
* kerberos
*/
u32 tmp[4];
- #ifdef VECT_SIZE1
-
tmp[0] = digest[0];
tmp[1] = digest[1];
tmp[2] = digest[2];
if (decrypt_and_check (&rc4_keys[lid], tmp, timestamp_ct) == 1)
{
- mark_hash (plains_buf, hashes_shown, digests_offset, gid, il_pos);
-
- d_return_buf[lid] = 1;
+ mark_hash (plains_buf, d_return_buf, salt_pos, 0, digests_offset + 0, gid, il_pos);
}
-
- #endif
}
}
-__kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m07500_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 krb5pa_t *krb5pa_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 combs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
+__kernel void m07500_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 krb5pa_t *krb5pa_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 il_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))) m07500_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 krb5pa_t *krb5pa_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 combs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
+__kernel void m07500_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 krb5pa_t *krb5pa_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 il_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))) m07500_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 krb5pa_t *krb5pa_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 combs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
+__kernel void m07500_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 krb5pa_t *krb5pa_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 il_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
{
/**
* modifier
*/
- __local RC4_KEY rc4_keys[64];
-
- const u32 gid = get_global_id (0);
const u32 lid = get_local_id (0);
- if (gid >= gid_max) return;
-
/**
* base
*/
- u32 wordl0[4];
-
- wordl0[0] = pws[gid].i[ 0];
- wordl0[1] = pws[gid].i[ 1];
- wordl0[2] = pws[gid].i[ 2];
- wordl0[3] = pws[gid].i[ 3];
-
- u32 wordl1[4];
-
- wordl1[0] = pws[gid].i[ 4];
- wordl1[1] = pws[gid].i[ 5];
- wordl1[2] = pws[gid].i[ 6];
- wordl1[3] = pws[gid].i[ 7];
-
- u32 wordl2[4];
+ const u32 gid = get_global_id (0);
- wordl2[0] = 0;
- wordl2[1] = 0;
- wordl2[2] = 0;
- wordl2[3] = 0;
+ if (gid >= gid_max) return;
- u32 wordl3[4];
+ u32 pw_buf0[4];
+ u32 pw_buf1[4];
- wordl3[0] = 0;
- wordl3[1] = 0;
- wordl3[2] = 0;
- wordl3[3] = 0;
+ 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];
+ 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_l_len = pws[gid].pw_len;
- if (combs_mode == COMBINATOR_MODE_BASE_RIGHT)
- {
- switch_buffer_by_offset (wordl0, wordl1, wordl2, wordl3, combs_buf[0].pw_len);
- }
-
/**
* salt
*/
timestamp_ct[7] = krb5pa_bufs[salt_pos].timestamp[7];
/**
- * loop
+ * shared
*/
- for (u32 il_pos = 0; il_pos < combs_cnt; il_pos++)
- {
- const u32 pw_r_len = combs_buf[il_pos].pw_len;
-
- const u32 pw_len = pw_l_len + pw_r_len;
-
- u32 wordr0[4];
-
- wordr0[0] = combs_buf[il_pos].i[0];
- wordr0[1] = combs_buf[il_pos].i[1];
- wordr0[2] = combs_buf[il_pos].i[2];
- wordr0[3] = combs_buf[il_pos].i[3];
-
- u32 wordr1[4];
+ __local RC4_KEY rc4_keys[64];
- wordr1[0] = combs_buf[il_pos].i[4];
- wordr1[1] = combs_buf[il_pos].i[5];
- wordr1[2] = combs_buf[il_pos].i[6];
- wordr1[3] = combs_buf[il_pos].i[7];
+ /**
+ * loop
+ */
- u32 wordr2[4];
+ for (u32 il_pos = 0; il_pos < il_cnt; il_pos += VECT_SIZE)
+ {
+ const u32x pw_r_len = pwlenx_create_combt (combs_buf, il_pos);
- wordr2[0] = 0;
- wordr2[1] = 0;
- wordr2[2] = 0;
- wordr2[3] = 0;
+ const u32x pw_len = pw_l_len + pw_r_len;
- u32 wordr3[4];
+ /**
+ * concat password candidate
+ */
- wordr3[0] = 0;
- wordr3[1] = 0;
- wordr3[2] = 0;
- wordr3[3] = 0;
+ u32x wordl0[4] = { 0 };
+ u32x wordl1[4] = { 0 };
+ u32x wordl2[4] = { 0 };
+ u32x wordl3[4] = { 0 };
+
+ wordl0[0] = pw_buf0[0];
+ wordl0[1] = pw_buf0[1];
+ wordl0[2] = pw_buf0[2];
+ wordl0[3] = pw_buf0[3];
+ wordl1[0] = pw_buf1[0];
+ wordl1[1] = pw_buf1[1];
+ wordl1[2] = pw_buf1[2];
+ wordl1[3] = pw_buf1[3];
+
+ u32x wordr0[4] = { 0 };
+ u32x wordr1[4] = { 0 };
+ u32x wordr2[4] = { 0 };
+ u32x wordr3[4] = { 0 };
+
+ wordr0[0] = ix_create_combt (combs_buf, il_pos, 0);
+ wordr0[1] = ix_create_combt (combs_buf, il_pos, 1);
+ wordr0[2] = ix_create_combt (combs_buf, il_pos, 2);
+ wordr0[3] = ix_create_combt (combs_buf, il_pos, 3);
+ wordr1[0] = ix_create_combt (combs_buf, il_pos, 4);
+ wordr1[1] = ix_create_combt (combs_buf, il_pos, 5);
+ wordr1[2] = ix_create_combt (combs_buf, il_pos, 6);
+ wordr1[3] = ix_create_combt (combs_buf, il_pos, 7);
if (combs_mode == COMBINATOR_MODE_BASE_LEFT)
{
- switch_buffer_by_offset (wordr0, wordr1, wordr2, wordr3, pw_l_len);
+ switch_buffer_by_offset_le_VV (wordr0, wordr1, wordr2, wordr3, pw_l_len);
+ }
+ else
+ {
+ switch_buffer_by_offset_le_VV (wordl0, wordl1, wordl2, wordl3, pw_r_len);
}
- u32 w0[4];
+ u32x w0[4];
+ u32x w1[4];
w0[0] = wordl0[0] | wordr0[0];
w0[1] = wordl0[1] | wordr0[1];
w0[2] = wordl0[2] | wordr0[2];
w0[3] = wordl0[3] | wordr0[3];
-
- u32 w1[4];
-
w1[0] = wordl1[0] | wordr1[0];
w1[1] = wordl1[1] | wordr1[1];
w1[2] = wordl1[2] | wordr1[2];
w1[3] = wordl1[3] | wordr1[3];
- u32 w2[4];
-
- w2[0] = wordl2[0] | wordr2[0];
- w2[1] = wordl2[1] | wordr2[1];
- w2[2] = wordl2[2] | wordr2[2];
- w2[3] = wordl2[3] | wordr2[3];
-
- u32 w3[4];
-
- w3[0] = wordl3[0] | wordr3[0];
- w3[1] = wordl3[1] | wordr3[1];
- w3[2] = 0;
- w3[3] = 0;
-
/**
* kerberos
*/
u32 tmp[4];
- #ifdef VECT_SIZE1
-
tmp[0] = digest[0];
tmp[1] = digest[1];
tmp[2] = digest[2];
if (decrypt_and_check (&rc4_keys[lid], tmp, timestamp_ct) == 1)
{
- mark_hash (plains_buf, hashes_shown, digests_offset, gid, il_pos);
-
- d_return_buf[lid] = 1;
+ mark_hash (plains_buf, d_return_buf, salt_pos, 0, digests_offset + 0, gid, il_pos);
}
-
- #endif
}
}
-__kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m07500_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 krb5pa_t *krb5pa_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 combs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
+__kernel void m07500_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 krb5pa_t *krb5pa_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 il_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))) m07500_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 krb5pa_t *krb5pa_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 combs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
+__kernel void m07500_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 krb5pa_t *krb5pa_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 il_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
{
}