2 * Author......: Jens Steube <jens.steube@gmail.com>
8 #include "include/constants.h"
9 #include "include/kernel_vendor.h"
16 #include "include/kernel_functions.c"
17 #include "OpenCL/types_ocl.c"
18 #include "OpenCL/common.c"
20 #define COMPARE_S "OpenCL/check_single_comp4.c"
21 #define COMPARE_M "OpenCL/check_multi_comp4.c"
31 static void swap (__local RC4_KEY *rc4_key, const u8 i, const u8 j)
36 rc4_key->S[i] = rc4_key->S[j];
40 static void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
45 __local u32 *ptr = (__local u32 *) rc4_key->S;
48 for (u32 i = 0; i < 64; i++)
55 for (u32 i = 0; i < 16; i++)
63 j += rc4_key->S[idx] + (v >> 0); swap (rc4_key, idx, j); idx++;
64 j += rc4_key->S[idx] + (v >> 8); swap (rc4_key, idx, j); idx++;
65 j += rc4_key->S[idx] + (v >> 16); swap (rc4_key, idx, j); idx++;
66 j += rc4_key->S[idx] + (v >> 24); swap (rc4_key, idx, j); idx++;
70 j += rc4_key->S[idx] + (v >> 0); swap (rc4_key, idx, j); idx++;
71 j += rc4_key->S[idx] + (v >> 8); swap (rc4_key, idx, j); idx++;
72 j += rc4_key->S[idx] + (v >> 16); swap (rc4_key, idx, j); idx++;
73 j += rc4_key->S[idx] + (v >> 24); swap (rc4_key, idx, j); idx++;
77 j += rc4_key->S[idx] + (v >> 0); swap (rc4_key, idx, j); idx++;
78 j += rc4_key->S[idx] + (v >> 8); swap (rc4_key, idx, j); idx++;
79 j += rc4_key->S[idx] + (v >> 16); swap (rc4_key, idx, j); idx++;
80 j += rc4_key->S[idx] + (v >> 24); swap (rc4_key, idx, j); idx++;
84 j += rc4_key->S[idx] + (v >> 0); swap (rc4_key, idx, j); idx++;
85 j += rc4_key->S[idx] + (v >> 8); swap (rc4_key, idx, j); idx++;
86 j += rc4_key->S[idx] + (v >> 16); swap (rc4_key, idx, j); idx++;
87 j += rc4_key->S[idx] + (v >> 24); swap (rc4_key, idx, j); idx++;
91 static u8 rc4_next_16 (__local RC4_KEY *rc4_key, u8 i, u8 j, const u32 in[4], u32 out[4])
94 for (u32 k = 0; k < 4; k++)
103 swap (rc4_key, i, j);
105 idx = rc4_key->S[i] + rc4_key->S[j];
107 xor4 |= rc4_key->S[idx] << 0;
112 swap (rc4_key, i, j);
114 idx = rc4_key->S[i] + rc4_key->S[j];
116 xor4 |= rc4_key->S[idx] << 8;
121 swap (rc4_key, i, j);
123 idx = rc4_key->S[i] + rc4_key->S[j];
125 xor4 |= rc4_key->S[idx] << 16;
130 swap (rc4_key, i, j);
132 idx = rc4_key->S[i] + rc4_key->S[j];
134 xor4 |= rc4_key->S[idx] << 24;
136 out[k] = in[k] ^ xor4;
142 static void md5_transform (const u32 w0[4], const u32 w1[4], const u32 w2[4], const u32 w3[4], u32 digest[4])
166 MD5_STEP (MD5_Fo, a, b, c, d, w0_t, MD5C00, MD5S00);
167 MD5_STEP (MD5_Fo, d, a, b, c, w1_t, MD5C01, MD5S01);
168 MD5_STEP (MD5_Fo, c, d, a, b, w2_t, MD5C02, MD5S02);
169 MD5_STEP (MD5_Fo, b, c, d, a, w3_t, MD5C03, MD5S03);
170 MD5_STEP (MD5_Fo, a, b, c, d, w4_t, MD5C04, MD5S00);
171 MD5_STEP (MD5_Fo, d, a, b, c, w5_t, MD5C05, MD5S01);
172 MD5_STEP (MD5_Fo, c, d, a, b, w6_t, MD5C06, MD5S02);
173 MD5_STEP (MD5_Fo, b, c, d, a, w7_t, MD5C07, MD5S03);
174 MD5_STEP (MD5_Fo, a, b, c, d, w8_t, MD5C08, MD5S00);
175 MD5_STEP (MD5_Fo, d, a, b, c, w9_t, MD5C09, MD5S01);
176 MD5_STEP (MD5_Fo, c, d, a, b, wa_t, MD5C0a, MD5S02);
177 MD5_STEP (MD5_Fo, b, c, d, a, wb_t, MD5C0b, MD5S03);
178 MD5_STEP (MD5_Fo, a, b, c, d, wc_t, MD5C0c, MD5S00);
179 MD5_STEP (MD5_Fo, d, a, b, c, wd_t, MD5C0d, MD5S01);
180 MD5_STEP (MD5_Fo, c, d, a, b, we_t, MD5C0e, MD5S02);
181 MD5_STEP (MD5_Fo, b, c, d, a, wf_t, MD5C0f, MD5S03);
183 MD5_STEP (MD5_Go, a, b, c, d, w1_t, MD5C10, MD5S10);
184 MD5_STEP (MD5_Go, d, a, b, c, w6_t, MD5C11, MD5S11);
185 MD5_STEP (MD5_Go, c, d, a, b, wb_t, MD5C12, MD5S12);
186 MD5_STEP (MD5_Go, b, c, d, a, w0_t, MD5C13, MD5S13);
187 MD5_STEP (MD5_Go, a, b, c, d, w5_t, MD5C14, MD5S10);
188 MD5_STEP (MD5_Go, d, a, b, c, wa_t, MD5C15, MD5S11);
189 MD5_STEP (MD5_Go, c, d, a, b, wf_t, MD5C16, MD5S12);
190 MD5_STEP (MD5_Go, b, c, d, a, w4_t, MD5C17, MD5S13);
191 MD5_STEP (MD5_Go, a, b, c, d, w9_t, MD5C18, MD5S10);
192 MD5_STEP (MD5_Go, d, a, b, c, we_t, MD5C19, MD5S11);
193 MD5_STEP (MD5_Go, c, d, a, b, w3_t, MD5C1a, MD5S12);
194 MD5_STEP (MD5_Go, b, c, d, a, w8_t, MD5C1b, MD5S13);
195 MD5_STEP (MD5_Go, a, b, c, d, wd_t, MD5C1c, MD5S10);
196 MD5_STEP (MD5_Go, d, a, b, c, w2_t, MD5C1d, MD5S11);
197 MD5_STEP (MD5_Go, c, d, a, b, w7_t, MD5C1e, MD5S12);
198 MD5_STEP (MD5_Go, b, c, d, a, wc_t, MD5C1f, MD5S13);
200 MD5_STEP (MD5_H , a, b, c, d, w5_t, MD5C20, MD5S20);
201 MD5_STEP (MD5_H , d, a, b, c, w8_t, MD5C21, MD5S21);
202 MD5_STEP (MD5_H , c, d, a, b, wb_t, MD5C22, MD5S22);
203 MD5_STEP (MD5_H , b, c, d, a, we_t, MD5C23, MD5S23);
204 MD5_STEP (MD5_H , a, b, c, d, w1_t, MD5C24, MD5S20);
205 MD5_STEP (MD5_H , d, a, b, c, w4_t, MD5C25, MD5S21);
206 MD5_STEP (MD5_H , c, d, a, b, w7_t, MD5C26, MD5S22);
207 MD5_STEP (MD5_H , b, c, d, a, wa_t, MD5C27, MD5S23);
208 MD5_STEP (MD5_H , a, b, c, d, wd_t, MD5C28, MD5S20);
209 MD5_STEP (MD5_H , d, a, b, c, w0_t, MD5C29, MD5S21);
210 MD5_STEP (MD5_H , c, d, a, b, w3_t, MD5C2a, MD5S22);
211 MD5_STEP (MD5_H , b, c, d, a, w6_t, MD5C2b, MD5S23);
212 MD5_STEP (MD5_H , a, b, c, d, w9_t, MD5C2c, MD5S20);
213 MD5_STEP (MD5_H , d, a, b, c, wc_t, MD5C2d, MD5S21);
214 MD5_STEP (MD5_H , c, d, a, b, wf_t, MD5C2e, MD5S22);
215 MD5_STEP (MD5_H , b, c, d, a, w2_t, MD5C2f, MD5S23);
217 MD5_STEP (MD5_I , a, b, c, d, w0_t, MD5C30, MD5S30);
218 MD5_STEP (MD5_I , d, a, b, c, w7_t, MD5C31, MD5S31);
219 MD5_STEP (MD5_I , c, d, a, b, we_t, MD5C32, MD5S32);
220 MD5_STEP (MD5_I , b, c, d, a, w5_t, MD5C33, MD5S33);
221 MD5_STEP (MD5_I , a, b, c, d, wc_t, MD5C34, MD5S30);
222 MD5_STEP (MD5_I , d, a, b, c, w3_t, MD5C35, MD5S31);
223 MD5_STEP (MD5_I , c, d, a, b, wa_t, MD5C36, MD5S32);
224 MD5_STEP (MD5_I , b, c, d, a, w1_t, MD5C37, MD5S33);
225 MD5_STEP (MD5_I , a, b, c, d, w8_t, MD5C38, MD5S30);
226 MD5_STEP (MD5_I , d, a, b, c, wf_t, MD5C39, MD5S31);
227 MD5_STEP (MD5_I , c, d, a, b, w6_t, MD5C3a, MD5S32);
228 MD5_STEP (MD5_I , b, c, d, a, wd_t, MD5C3b, MD5S33);
229 MD5_STEP (MD5_I , a, b, c, d, w4_t, MD5C3c, MD5S30);
230 MD5_STEP (MD5_I , d, a, b, c, wb_t, MD5C3d, MD5S31);
231 MD5_STEP (MD5_I , c, d, a, b, w2_t, MD5C3e, MD5S32);
232 MD5_STEP (MD5_I , b, c, d, a, w9_t, MD5C3f, MD5S33);
240 __kernel void m09710_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 oldoffice01_t *oldoffice01_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)
246 const u32 lid = get_local_id (0);
248 __local RC4_KEY rc4_keys[64];
250 __local RC4_KEY *rc4_key = &rc4_keys[lid];
256 const u32 gid = get_global_id (0);
258 if (gid >= gid_max) return;
262 wordl0[0] = pws[gid].i[ 0];
263 wordl0[1] = pws[gid].i[ 1];
264 wordl0[2] = pws[gid].i[ 2];
265 wordl0[3] = pws[gid].i[ 3];
269 wordl1[0] = pws[gid].i[ 4];
270 wordl1[1] = pws[gid].i[ 5];
271 wordl1[2] = pws[gid].i[ 6];
272 wordl1[3] = pws[gid].i[ 7];
288 const u32 pw_l_len = pws[gid].pw_len;
290 if (combs_mode == COMBINATOR_MODE_BASE_RIGHT)
292 switch_buffer_by_offset_le (wordl0, wordl1, wordl2, wordl3, combs_buf[0].pw_len);
299 const u32 version = oldoffice01_bufs[salt_pos].version;
301 u32 encryptedVerifier[4];
303 encryptedVerifier[0] = oldoffice01_bufs[salt_pos].encryptedVerifier[0];
304 encryptedVerifier[1] = oldoffice01_bufs[salt_pos].encryptedVerifier[1];
305 encryptedVerifier[2] = oldoffice01_bufs[salt_pos].encryptedVerifier[2];
306 encryptedVerifier[3] = oldoffice01_bufs[salt_pos].encryptedVerifier[3];
312 for (u32 il_pos = 0; il_pos < il_cnt; il_pos++)
314 const u32 pw_r_len = combs_buf[il_pos].pw_len;
316 const u32 pw_len = pw_l_len + pw_r_len;
320 wordr0[0] = combs_buf[il_pos].i[0];
321 wordr0[1] = combs_buf[il_pos].i[1];
322 wordr0[2] = combs_buf[il_pos].i[2];
323 wordr0[3] = combs_buf[il_pos].i[3];
327 wordr1[0] = combs_buf[il_pos].i[4];
328 wordr1[1] = combs_buf[il_pos].i[5];
329 wordr1[2] = combs_buf[il_pos].i[6];
330 wordr1[3] = combs_buf[il_pos].i[7];
346 if (combs_mode == COMBINATOR_MODE_BASE_LEFT)
348 switch_buffer_by_offset_le (wordr0, wordr1, wordr2, wordr3, pw_l_len);
351 // first md5 to generate RC4 128 bit key
358 w0_t[0] = wordl0[0] | wordr0[0];
359 w0_t[1] = (wordl0[1] | wordr0[1]) & 0xff;
382 md5_transform (w0_t, w1_t, w2_t, w3_t, digest);
393 rc4_init_16 (rc4_key, key);
397 u8 j = rc4_next_16 (rc4_key, 0, 0, encryptedVerifier, out);
421 md5_transform (w0_t, w1_t, w2_t, w3_t, digest);
423 rc4_next_16 (rc4_key, 16, j, digest, out);
425 const u32 r0 = out[0];
426 const u32 r1 = out[1];
427 const u32 r2 = out[2];
428 const u32 r3 = out[3];
434 __kernel void m09710_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 oldoffice01_t *oldoffice01_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)
438 __kernel void m09710_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 oldoffice01_t *oldoffice01_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)
442 __kernel void m09710_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 oldoffice01_t *oldoffice01_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)
448 const u32 lid = get_local_id (0);
450 __local RC4_KEY rc4_keys[64];
452 __local RC4_KEY *rc4_key = &rc4_keys[lid];
458 const u32 gid = get_global_id (0);
460 if (gid >= gid_max) return;
464 wordl0[0] = pws[gid].i[ 0];
465 wordl0[1] = pws[gid].i[ 1];
466 wordl0[2] = pws[gid].i[ 2];
467 wordl0[3] = pws[gid].i[ 3];
471 wordl1[0] = pws[gid].i[ 4];
472 wordl1[1] = pws[gid].i[ 5];
473 wordl1[2] = pws[gid].i[ 6];
474 wordl1[3] = pws[gid].i[ 7];
490 const u32 pw_l_len = pws[gid].pw_len;
492 if (combs_mode == COMBINATOR_MODE_BASE_RIGHT)
494 switch_buffer_by_offset_le (wordl0, wordl1, wordl2, wordl3, combs_buf[0].pw_len);
501 const u32 search[4] =
503 digests_buf[digests_offset].digest_buf[DGST_R0],
504 digests_buf[digests_offset].digest_buf[DGST_R1],
505 digests_buf[digests_offset].digest_buf[DGST_R2],
506 digests_buf[digests_offset].digest_buf[DGST_R3]
513 const u32 version = oldoffice01_bufs[salt_pos].version;
515 u32 encryptedVerifier[4];
517 encryptedVerifier[0] = oldoffice01_bufs[salt_pos].encryptedVerifier[0];
518 encryptedVerifier[1] = oldoffice01_bufs[salt_pos].encryptedVerifier[1];
519 encryptedVerifier[2] = oldoffice01_bufs[salt_pos].encryptedVerifier[2];
520 encryptedVerifier[3] = oldoffice01_bufs[salt_pos].encryptedVerifier[3];
526 for (u32 il_pos = 0; il_pos < il_cnt; il_pos++)
528 const u32 pw_r_len = combs_buf[il_pos].pw_len;
530 const u32 pw_len = pw_l_len + pw_r_len;
534 wordr0[0] = combs_buf[il_pos].i[0];
535 wordr0[1] = combs_buf[il_pos].i[1];
536 wordr0[2] = combs_buf[il_pos].i[2];
537 wordr0[3] = combs_buf[il_pos].i[3];
541 wordr1[0] = combs_buf[il_pos].i[4];
542 wordr1[1] = combs_buf[il_pos].i[5];
543 wordr1[2] = combs_buf[il_pos].i[6];
544 wordr1[3] = combs_buf[il_pos].i[7];
560 if (combs_mode == COMBINATOR_MODE_BASE_LEFT)
562 switch_buffer_by_offset_le (wordr0, wordr1, wordr2, wordr3, pw_l_len);
565 // first md5 to generate RC4 128 bit key
572 w0_t[0] = wordl0[0] | wordr0[0];
573 w0_t[1] = (wordl0[1] | wordr0[1]) & 0xff;
596 md5_transform (w0_t, w1_t, w2_t, w3_t, digest);
607 rc4_init_16 (rc4_key, key);
611 u8 j = rc4_next_16 (rc4_key, 0, 0, encryptedVerifier, out);
635 md5_transform (w0_t, w1_t, w2_t, w3_t, digest);
637 rc4_next_16 (rc4_key, 16, j, digest, out);
639 const u32 r0 = out[0];
640 const u32 r1 = out[1];
641 const u32 r2 = out[2];
642 const u32 r3 = out[3];
648 __kernel void m09710_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 oldoffice01_t *oldoffice01_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)
652 __kernel void m09710_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 oldoffice01_t *oldoffice01_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)