2 * Author......: Jens Steube <jens.steube@gmail.com>
8 #include "include/constants.h"
9 #include "include/kernel_vendor.h"
28 #include "include/kernel_functions.c"
29 #include "types_amd.c"
30 #include "common_amd.c"
31 #include "include/rp_gpu.h"
35 #define VECT_COMPARE_S "check_single_vect1_comp4.c"
36 #define VECT_COMPARE_M "check_multi_vect1_comp4.c"
40 #define VECT_COMPARE_S "check_single_vect2_comp4.c"
41 #define VECT_COMPARE_M "check_multi_vect2_comp4.c"
45 #define VECT_COMPARE_S "check_single_vect4_comp4.c"
46 #define VECT_COMPARE_M "check_multi_vect4_comp4.c"
57 static void swap (__local RC4_KEY *rc4_key, const u8 i, const u8 j)
62 rc4_key->S[i] = rc4_key->S[j];
66 static void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
71 __local u32 *ptr = (__local u32 *) rc4_key->S;
74 for (u32 i = 0; i < 64; i++)
82 for (u32 i = 0; i < 16; i++)
90 j += rc4_key->S[idx] + (v >> 0); swap (rc4_key, idx, j); idx++;
91 j += rc4_key->S[idx] + (v >> 8); swap (rc4_key, idx, j); idx++;
92 j += rc4_key->S[idx] + (v >> 16); swap (rc4_key, idx, j); idx++;
93 j += rc4_key->S[idx] + (v >> 24); swap (rc4_key, idx, j); idx++;
97 j += rc4_key->S[idx] + (v >> 0); swap (rc4_key, idx, j); idx++;
98 j += rc4_key->S[idx] + (v >> 8); swap (rc4_key, idx, j); idx++;
99 j += rc4_key->S[idx] + (v >> 16); swap (rc4_key, idx, j); idx++;
100 j += rc4_key->S[idx] + (v >> 24); swap (rc4_key, idx, j); idx++;
104 j += rc4_key->S[idx] + (v >> 0); swap (rc4_key, idx, j); idx++;
105 j += rc4_key->S[idx] + (v >> 8); swap (rc4_key, idx, j); idx++;
106 j += rc4_key->S[idx] + (v >> 16); swap (rc4_key, idx, j); idx++;
107 j += rc4_key->S[idx] + (v >> 24); swap (rc4_key, idx, j); idx++;
111 j += rc4_key->S[idx] + (v >> 0); swap (rc4_key, idx, j); idx++;
112 j += rc4_key->S[idx] + (v >> 8); swap (rc4_key, idx, j); idx++;
113 j += rc4_key->S[idx] + (v >> 16); swap (rc4_key, idx, j); idx++;
114 j += rc4_key->S[idx] + (v >> 24); swap (rc4_key, idx, j); idx++;
118 static u8 rc4_next_16 (__local RC4_KEY *rc4_key, u8 i, u8 j, const u32 in[4], u32 out[4])
121 for (u32 k = 0; k < 4; k++)
130 swap (rc4_key, i, j);
132 idx = rc4_key->S[i] + rc4_key->S[j];
134 xor4 |= rc4_key->S[idx] << 0;
139 swap (rc4_key, i, j);
141 idx = rc4_key->S[i] + rc4_key->S[j];
143 xor4 |= rc4_key->S[idx] << 8;
148 swap (rc4_key, i, j);
150 idx = rc4_key->S[i] + rc4_key->S[j];
152 xor4 |= rc4_key->S[idx] << 16;
157 swap (rc4_key, i, j);
159 idx = rc4_key->S[i] + rc4_key->S[j];
161 xor4 |= rc4_key->S[idx] << 24;
163 out[k] = in[k] ^ xor4;
169 static void md5_transform (const u32x w0[4], const u32x w1[4], const u32x w2[4], const u32x w3[4], u32x digest[4])
193 MD5_STEP (MD5_Fo, a, b, c, d, w0_t, MD5C00, MD5S00);
194 MD5_STEP (MD5_Fo, d, a, b, c, w1_t, MD5C01, MD5S01);
195 MD5_STEP (MD5_Fo, c, d, a, b, w2_t, MD5C02, MD5S02);
196 MD5_STEP (MD5_Fo, b, c, d, a, w3_t, MD5C03, MD5S03);
197 MD5_STEP (MD5_Fo, a, b, c, d, w4_t, MD5C04, MD5S00);
198 MD5_STEP (MD5_Fo, d, a, b, c, w5_t, MD5C05, MD5S01);
199 MD5_STEP (MD5_Fo, c, d, a, b, w6_t, MD5C06, MD5S02);
200 MD5_STEP (MD5_Fo, b, c, d, a, w7_t, MD5C07, MD5S03);
201 MD5_STEP (MD5_Fo, a, b, c, d, w8_t, MD5C08, MD5S00);
202 MD5_STEP (MD5_Fo, d, a, b, c, w9_t, MD5C09, MD5S01);
203 MD5_STEP (MD5_Fo, c, d, a, b, wa_t, MD5C0a, MD5S02);
204 MD5_STEP (MD5_Fo, b, c, d, a, wb_t, MD5C0b, MD5S03);
205 MD5_STEP (MD5_Fo, a, b, c, d, wc_t, MD5C0c, MD5S00);
206 MD5_STEP (MD5_Fo, d, a, b, c, wd_t, MD5C0d, MD5S01);
207 MD5_STEP (MD5_Fo, c, d, a, b, we_t, MD5C0e, MD5S02);
208 MD5_STEP (MD5_Fo, b, c, d, a, wf_t, MD5C0f, MD5S03);
210 MD5_STEP (MD5_Go, a, b, c, d, w1_t, MD5C10, MD5S10);
211 MD5_STEP (MD5_Go, d, a, b, c, w6_t, MD5C11, MD5S11);
212 MD5_STEP (MD5_Go, c, d, a, b, wb_t, MD5C12, MD5S12);
213 MD5_STEP (MD5_Go, b, c, d, a, w0_t, MD5C13, MD5S13);
214 MD5_STEP (MD5_Go, a, b, c, d, w5_t, MD5C14, MD5S10);
215 MD5_STEP (MD5_Go, d, a, b, c, wa_t, MD5C15, MD5S11);
216 MD5_STEP (MD5_Go, c, d, a, b, wf_t, MD5C16, MD5S12);
217 MD5_STEP (MD5_Go, b, c, d, a, w4_t, MD5C17, MD5S13);
218 MD5_STEP (MD5_Go, a, b, c, d, w9_t, MD5C18, MD5S10);
219 MD5_STEP (MD5_Go, d, a, b, c, we_t, MD5C19, MD5S11);
220 MD5_STEP (MD5_Go, c, d, a, b, w3_t, MD5C1a, MD5S12);
221 MD5_STEP (MD5_Go, b, c, d, a, w8_t, MD5C1b, MD5S13);
222 MD5_STEP (MD5_Go, a, b, c, d, wd_t, MD5C1c, MD5S10);
223 MD5_STEP (MD5_Go, d, a, b, c, w2_t, MD5C1d, MD5S11);
224 MD5_STEP (MD5_Go, c, d, a, b, w7_t, MD5C1e, MD5S12);
225 MD5_STEP (MD5_Go, b, c, d, a, wc_t, MD5C1f, MD5S13);
227 MD5_STEP (MD5_H , a, b, c, d, w5_t, MD5C20, MD5S20);
228 MD5_STEP (MD5_H , d, a, b, c, w8_t, MD5C21, MD5S21);
229 MD5_STEP (MD5_H , c, d, a, b, wb_t, MD5C22, MD5S22);
230 MD5_STEP (MD5_H , b, c, d, a, we_t, MD5C23, MD5S23);
231 MD5_STEP (MD5_H , a, b, c, d, w1_t, MD5C24, MD5S20);
232 MD5_STEP (MD5_H , d, a, b, c, w4_t, MD5C25, MD5S21);
233 MD5_STEP (MD5_H , c, d, a, b, w7_t, MD5C26, MD5S22);
234 MD5_STEP (MD5_H , b, c, d, a, wa_t, MD5C27, MD5S23);
235 MD5_STEP (MD5_H , a, b, c, d, wd_t, MD5C28, MD5S20);
236 MD5_STEP (MD5_H , d, a, b, c, w0_t, MD5C29, MD5S21);
237 MD5_STEP (MD5_H , c, d, a, b, w3_t, MD5C2a, MD5S22);
238 MD5_STEP (MD5_H , b, c, d, a, w6_t, MD5C2b, MD5S23);
239 MD5_STEP (MD5_H , a, b, c, d, w9_t, MD5C2c, MD5S20);
240 MD5_STEP (MD5_H , d, a, b, c, wc_t, MD5C2d, MD5S21);
241 MD5_STEP (MD5_H , c, d, a, b, wf_t, MD5C2e, MD5S22);
242 MD5_STEP (MD5_H , b, c, d, a, w2_t, MD5C2f, MD5S23);
244 MD5_STEP (MD5_I , a, b, c, d, w0_t, MD5C30, MD5S30);
245 MD5_STEP (MD5_I , d, a, b, c, w7_t, MD5C31, MD5S31);
246 MD5_STEP (MD5_I , c, d, a, b, we_t, MD5C32, MD5S32);
247 MD5_STEP (MD5_I , b, c, d, a, w5_t, MD5C33, MD5S33);
248 MD5_STEP (MD5_I , a, b, c, d, wc_t, MD5C34, MD5S30);
249 MD5_STEP (MD5_I , d, a, b, c, w3_t, MD5C35, MD5S31);
250 MD5_STEP (MD5_I , c, d, a, b, wa_t, MD5C36, MD5S32);
251 MD5_STEP (MD5_I , b, c, d, a, w1_t, MD5C37, MD5S33);
252 MD5_STEP (MD5_I , a, b, c, d, w8_t, MD5C38, MD5S30);
253 MD5_STEP (MD5_I , d, a, b, c, wf_t, MD5C39, MD5S31);
254 MD5_STEP (MD5_I , c, d, a, b, w6_t, MD5C3a, MD5S32);
255 MD5_STEP (MD5_I , b, c, d, a, wd_t, MD5C3b, MD5S33);
256 MD5_STEP (MD5_I , a, b, c, d, w4_t, MD5C3c, MD5S30);
257 MD5_STEP (MD5_I , d, a, b, c, wb_t, MD5C3d, MD5S31);
258 MD5_STEP (MD5_I , c, d, a, b, w2_t, MD5C3e, MD5S32);
259 MD5_STEP (MD5_I , b, c, d, a, w9_t, MD5C3f, MD5S33);
267 static void gen336 (u32x digest_pre[4], u32 salt_buf[4], u32x digest[4])
274 digest_t0[0] = digest_pre[0];
275 digest_t0[1] = digest_pre[1] & 0xff;
277 digest_t1[0] = digest_pre[0] << 8;
278 digest_t1[1] = digest_pre[0] >> 24 | digest_pre[1] << 8;
280 digest_t2[0] = digest_pre[0] << 16;
281 digest_t2[1] = digest_pre[0] >> 16 | digest_pre[1] << 16;
283 digest_t3[0] = digest_pre[0] << 24;
284 digest_t3[1] = digest_pre[0] >> 8 | digest_pre[1] << 24;
291 salt_buf_t0[0] = salt_buf[0];
292 salt_buf_t0[1] = salt_buf[1];
293 salt_buf_t0[2] = salt_buf[2];
294 salt_buf_t0[3] = salt_buf[3];
296 salt_buf_t1[0] = salt_buf[0] << 8;
297 salt_buf_t1[1] = salt_buf[0] >> 24 | salt_buf[1] << 8;
298 salt_buf_t1[2] = salt_buf[1] >> 24 | salt_buf[2] << 8;
299 salt_buf_t1[3] = salt_buf[2] >> 24 | salt_buf[3] << 8;
300 salt_buf_t1[4] = salt_buf[3] >> 24;
302 salt_buf_t2[0] = salt_buf[0] << 16;
303 salt_buf_t2[1] = salt_buf[0] >> 16 | salt_buf[1] << 16;
304 salt_buf_t2[2] = salt_buf[1] >> 16 | salt_buf[2] << 16;
305 salt_buf_t2[3] = salt_buf[2] >> 16 | salt_buf[3] << 16;
306 salt_buf_t2[4] = salt_buf[3] >> 16;
308 salt_buf_t3[0] = salt_buf[0] << 24;
309 salt_buf_t3[1] = salt_buf[0] >> 8 | salt_buf[1] << 24;
310 salt_buf_t3[2] = salt_buf[1] >> 8 | salt_buf[2] << 24;
311 salt_buf_t3[3] = salt_buf[2] >> 8 | salt_buf[3] << 24;
312 salt_buf_t3[4] = salt_buf[3] >> 8;
319 // generate the 16 * 21 buffer
339 w0_t[0] = digest_t0[0];
340 w0_t[1] = digest_t0[1];
343 w0_t[1] |= salt_buf_t1[0];
344 w0_t[2] = salt_buf_t1[1];
345 w0_t[3] = salt_buf_t1[2];
346 w1_t[0] = salt_buf_t1[3];
347 w1_t[1] = salt_buf_t1[4];
350 w1_t[1] |= digest_t1[0];
351 w1_t[2] = digest_t1[1];
354 w1_t[2] |= salt_buf_t2[0];
355 w1_t[3] = salt_buf_t2[1];
356 w2_t[0] = salt_buf_t2[2];
357 w2_t[1] = salt_buf_t2[3];
358 w2_t[2] = salt_buf_t2[4];
361 w2_t[2] |= digest_t2[0];
362 w2_t[3] = digest_t2[1];
365 w2_t[3] |= salt_buf_t3[0];
366 w3_t[0] = salt_buf_t3[1];
367 w3_t[1] = salt_buf_t3[2];
368 w3_t[2] = salt_buf_t3[3];
369 w3_t[3] = salt_buf_t3[4];
373 w3_t[3] |= digest_t3[0];
375 md5_transform (w0_t, w1_t, w2_t, w3_t, digest);
395 w0_t[0] = digest_t3[1];
398 w0_t[1] = salt_buf_t0[0];
399 w0_t[2] = salt_buf_t0[1];
400 w0_t[3] = salt_buf_t0[2];
401 w1_t[0] = salt_buf_t0[3];
404 w1_t[1] = digest_t0[0];
405 w1_t[2] = digest_t0[1];
408 w1_t[2] |= salt_buf_t1[0];
409 w1_t[3] = salt_buf_t1[1];
410 w2_t[0] = salt_buf_t1[2];
411 w2_t[1] = salt_buf_t1[3];
412 w2_t[2] = salt_buf_t1[4];
415 w2_t[2] |= digest_t1[0];
416 w2_t[3] = digest_t1[1];
419 w2_t[3] |= salt_buf_t2[0];
420 w3_t[0] = salt_buf_t2[1];
421 w3_t[1] = salt_buf_t2[2];
422 w3_t[2] = salt_buf_t2[3];
423 w3_t[3] = salt_buf_t2[4];
426 w3_t[3] |= digest_t2[0];
428 md5_transform (w0_t, w1_t, w2_t, w3_t, digest);
448 w0_t[0] = digest_t2[1];
451 w0_t[0] |= salt_buf_t3[0];
452 w0_t[1] = salt_buf_t3[1];
453 w0_t[2] = salt_buf_t3[2];
454 w0_t[3] = salt_buf_t3[3];
455 w1_t[0] = salt_buf_t3[4];
458 w1_t[0] |= digest_t3[0];
459 w1_t[1] = digest_t3[1];
462 w1_t[2] = salt_buf_t0[0];
463 w1_t[3] = salt_buf_t0[1];
464 w2_t[0] = salt_buf_t0[2];
465 w2_t[1] = salt_buf_t0[3];
468 w2_t[2] = digest_t0[0];
469 w2_t[3] = digest_t0[1];
472 w2_t[3] |= salt_buf_t1[0];
473 w3_t[0] = salt_buf_t1[1];
474 w3_t[1] = salt_buf_t1[2];
475 w3_t[2] = salt_buf_t1[3];
476 w3_t[3] = salt_buf_t1[4];
479 w3_t[3] |= digest_t1[0];
481 md5_transform (w0_t, w1_t, w2_t, w3_t, digest);
501 w0_t[0] = digest_t1[1];
504 w0_t[0] |= salt_buf_t2[0];
505 w0_t[1] = salt_buf_t2[1];
506 w0_t[2] = salt_buf_t2[2];
507 w0_t[3] = salt_buf_t2[3];
508 w1_t[0] = salt_buf_t2[4];
511 w1_t[0] |= digest_t2[0];
512 w1_t[1] = digest_t2[1];
515 w1_t[1] |= salt_buf_t3[0];
516 w1_t[2] = salt_buf_t3[1];
517 w1_t[3] = salt_buf_t3[2];
518 w2_t[0] = salt_buf_t3[3];
519 w2_t[1] = salt_buf_t3[4];
522 w2_t[1] |= digest_t3[0];
523 w2_t[2] = digest_t3[1];
526 w2_t[3] = salt_buf_t0[0];
527 w3_t[0] = salt_buf_t0[1];
528 w3_t[1] = salt_buf_t0[2];
529 w3_t[2] = salt_buf_t0[3];
532 w3_t[3] = digest_t0[0];
534 md5_transform (w0_t, w1_t, w2_t, w3_t, digest);
554 w0_t[0] = digest_t0[1];
557 w0_t[0] |= salt_buf_t1[0];
558 w0_t[1] = salt_buf_t1[1];
559 w0_t[2] = salt_buf_t1[2];
560 w0_t[3] = salt_buf_t1[3];
561 w1_t[0] = salt_buf_t1[4];
564 w1_t[0] |= digest_t1[0];
565 w1_t[1] = digest_t1[1];
568 w1_t[1] |= salt_buf_t2[0];
569 w1_t[2] = salt_buf_t2[1];
570 w1_t[3] = salt_buf_t2[2];
571 w2_t[0] = salt_buf_t2[3];
572 w2_t[1] = salt_buf_t2[4];
575 w2_t[1] |= digest_t2[0];
576 w2_t[2] = digest_t2[1];
579 w2_t[2] |= salt_buf_t3[0];
580 w2_t[3] = salt_buf_t3[1];
581 w3_t[0] = salt_buf_t3[2];
582 w3_t[1] = salt_buf_t3[3];
583 w3_t[2] = salt_buf_t3[4];
586 w3_t[2] |= digest_t3[0];
587 w3_t[3] = digest_t3[1];
589 md5_transform (w0_t, w1_t, w2_t, w3_t, digest);
591 w0_t[0] = salt_buf_t0[0];
592 w0_t[1] = salt_buf_t0[1];
593 w0_t[2] = salt_buf_t0[2];
594 w0_t[3] = salt_buf_t0[3];
605 w3_t[2] = 21 * 16 * 8;
608 md5_transform (w0_t, w1_t, w2_t, w3_t, digest);
611 __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m09700_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 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 rules_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
617 const u32 lid = get_local_id (0);
619 __local RC4_KEY rc4_keys[64];
621 __local RC4_KEY *rc4_key = &rc4_keys[lid];
627 const u32 gid = get_global_id (0);
629 if (gid >= gid_max) return;
633 pw_buf0[0] = pws[gid].i[ 0];
634 pw_buf0[1] = pws[gid].i[ 1];
635 pw_buf0[2] = pws[gid].i[ 2];
636 pw_buf0[3] = pws[gid].i[ 3];
640 pw_buf1[0] = pws[gid].i[ 4];
641 pw_buf1[1] = pws[gid].i[ 5];
642 pw_buf1[2] = pws[gid].i[ 6];
643 pw_buf1[3] = pws[gid].i[ 7];
645 const u32 pw_len = pws[gid].pw_len;
653 salt_buf[0] = salt_bufs[salt_pos].salt_buf[0];
654 salt_buf[1] = salt_bufs[salt_pos].salt_buf[1];
655 salt_buf[2] = salt_bufs[salt_pos].salt_buf[2];
656 salt_buf[3] = salt_bufs[salt_pos].salt_buf[3];
662 const u32 version = oldoffice01_bufs[salt_pos].version;
664 u32 encryptedVerifier[4];
666 encryptedVerifier[0] = oldoffice01_bufs[salt_pos].encryptedVerifier[0];
667 encryptedVerifier[1] = oldoffice01_bufs[salt_pos].encryptedVerifier[1];
668 encryptedVerifier[2] = oldoffice01_bufs[salt_pos].encryptedVerifier[2];
669 encryptedVerifier[3] = oldoffice01_bufs[salt_pos].encryptedVerifier[3];
675 for (u32 il_pos = 0; il_pos < rules_cnt; il_pos++)
705 const u32 out_len = apply_rules (rules_buf[il_pos].cmds, w0, w1, pw_len);
707 append_0x80_2 (w0, w1, out_len);
714 make_unicode (w0, w0_t, w1_t);
715 make_unicode (w1, w2_t, w3_t);
717 w3_t[2] = out_len * 8 * 2;
721 digest_pre[0] = MD5M_A;
722 digest_pre[1] = MD5M_B;
723 digest_pre[2] = MD5M_C;
724 digest_pre[3] = MD5M_D;
726 md5_transform (w0_t, w1_t, w2_t, w3_t, digest_pre);
728 digest_pre[0] &= 0xffffffff;
729 digest_pre[1] &= 0x000000ff;
730 digest_pre[2] &= 0x00000000;
731 digest_pre[3] &= 0x00000000;
740 gen336 (digest_pre, salt_buf, digest);
742 // now the 40 bit input for the MD5 which then will generate the RC4 key, so it's precomputable!
745 w0_t[1] = digest[1] & 0xff;
766 md5_transform (w0_t, w1_t, w2_t, w3_t, digest);
777 rc4_init_16 (rc4_key, key);
781 u8 j = rc4_next_16 (rc4_key, 0, 0, encryptedVerifier, out);
805 md5_transform (w0_t, w1_t, w2_t, w3_t, digest);
807 rc4_next_16 (rc4_key, 16, j, digest, out);
809 const u32x r0 = out[0];
810 const u32x r1 = out[1];
811 const u32x r2 = out[2];
812 const u32x r3 = out[3];
814 #include VECT_COMPARE_M
818 __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m09700_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 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 combs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
822 __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m09700_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 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 combs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
826 __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m09700_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 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 rules_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
832 const u32 lid = get_local_id (0);
834 __local RC4_KEY rc4_keys[64];
836 __local RC4_KEY *rc4_key = &rc4_keys[lid];
842 const u32 gid = get_global_id (0);
844 if (gid >= gid_max) return;
848 pw_buf0[0] = pws[gid].i[ 0];
849 pw_buf0[1] = pws[gid].i[ 1];
850 pw_buf0[2] = pws[gid].i[ 2];
851 pw_buf0[3] = pws[gid].i[ 3];
855 pw_buf1[0] = pws[gid].i[ 4];
856 pw_buf1[1] = pws[gid].i[ 5];
857 pw_buf1[2] = pws[gid].i[ 6];
858 pw_buf1[3] = pws[gid].i[ 7];
860 const u32 pw_len = pws[gid].pw_len;
866 const u32 search[4] =
868 digests_buf[digests_offset].digest_buf[DGST_R0],
869 digests_buf[digests_offset].digest_buf[DGST_R1],
870 digests_buf[digests_offset].digest_buf[DGST_R2],
871 digests_buf[digests_offset].digest_buf[DGST_R3]
880 salt_buf[0] = salt_bufs[salt_pos].salt_buf[0];
881 salt_buf[1] = salt_bufs[salt_pos].salt_buf[1];
882 salt_buf[2] = salt_bufs[salt_pos].salt_buf[2];
883 salt_buf[3] = salt_bufs[salt_pos].salt_buf[3];
889 const u32 version = oldoffice01_bufs[salt_pos].version;
891 u32 encryptedVerifier[4];
893 encryptedVerifier[0] = oldoffice01_bufs[salt_pos].encryptedVerifier[0];
894 encryptedVerifier[1] = oldoffice01_bufs[salt_pos].encryptedVerifier[1];
895 encryptedVerifier[2] = oldoffice01_bufs[salt_pos].encryptedVerifier[2];
896 encryptedVerifier[3] = oldoffice01_bufs[salt_pos].encryptedVerifier[3];
902 for (u32 il_pos = 0; il_pos < rules_cnt; il_pos++)
932 const u32 out_len = apply_rules (rules_buf[il_pos].cmds, w0, w1, pw_len);
934 append_0x80_2 (w0, w1, out_len);
941 make_unicode (w0, w0_t, w1_t);
942 make_unicode (w1, w2_t, w3_t);
944 w3_t[2] = out_len * 8 * 2;
948 digest_pre[0] = MD5M_A;
949 digest_pre[1] = MD5M_B;
950 digest_pre[2] = MD5M_C;
951 digest_pre[3] = MD5M_D;
953 md5_transform (w0_t, w1_t, w2_t, w3_t, digest_pre);
955 digest_pre[0] &= 0xffffffff;
956 digest_pre[1] &= 0x000000ff;
957 digest_pre[2] &= 0x00000000;
958 digest_pre[3] &= 0x00000000;
967 gen336 (digest_pre, salt_buf, digest);
969 // now the 40 bit input for the MD5 which then will generate the RC4 key, so it's precomputable!
972 w0_t[1] = digest[1] & 0xff;
993 md5_transform (w0_t, w1_t, w2_t, w3_t, digest);
1004 rc4_init_16 (rc4_key, key);
1008 u8 j = rc4_next_16 (rc4_key, 0, 0, encryptedVerifier, out);
1032 md5_transform (w0_t, w1_t, w2_t, w3_t, digest);
1034 rc4_next_16 (rc4_key, 16, j, digest, out);
1036 const u32x r0 = out[0];
1037 const u32x r1 = out[1];
1038 const u32x r2 = out[2];
1039 const u32x r3 = out[3];
1041 #include VECT_COMPARE_S
1045 __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m09700_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 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 combs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
1049 __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m09700_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 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 combs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)