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"
33 #define VECT_COMPARE_S "check_single_vect1_comp4.c"
34 #define VECT_COMPARE_M "check_multi_vect1_comp4.c"
38 #define VECT_COMPARE_S "check_single_vect2_comp4.c"
39 #define VECT_COMPARE_M "check_multi_vect2_comp4.c"
43 #define VECT_COMPARE_S "check_single_vect4_comp4.c"
44 #define VECT_COMPARE_M "check_multi_vect4_comp4.c"
47 __constant u32 lotus_magic_table[256] =
49 0xbd, 0x56, 0xea, 0xf2, 0xa2, 0xf1, 0xac, 0x2a,
50 0xb0, 0x93, 0xd1, 0x9c, 0x1b, 0x33, 0xfd, 0xd0,
51 0x30, 0x04, 0xb6, 0xdc, 0x7d, 0xdf, 0x32, 0x4b,
52 0xf7, 0xcb, 0x45, 0x9b, 0x31, 0xbb, 0x21, 0x5a,
53 0x41, 0x9f, 0xe1, 0xd9, 0x4a, 0x4d, 0x9e, 0xda,
54 0xa0, 0x68, 0x2c, 0xc3, 0x27, 0x5f, 0x80, 0x36,
55 0x3e, 0xee, 0xfb, 0x95, 0x1a, 0xfe, 0xce, 0xa8,
56 0x34, 0xa9, 0x13, 0xf0, 0xa6, 0x3f, 0xd8, 0x0c,
57 0x78, 0x24, 0xaf, 0x23, 0x52, 0xc1, 0x67, 0x17,
58 0xf5, 0x66, 0x90, 0xe7, 0xe8, 0x07, 0xb8, 0x60,
59 0x48, 0xe6, 0x1e, 0x53, 0xf3, 0x92, 0xa4, 0x72,
60 0x8c, 0x08, 0x15, 0x6e, 0x86, 0x00, 0x84, 0xfa,
61 0xf4, 0x7f, 0x8a, 0x42, 0x19, 0xf6, 0xdb, 0xcd,
62 0x14, 0x8d, 0x50, 0x12, 0xba, 0x3c, 0x06, 0x4e,
63 0xec, 0xb3, 0x35, 0x11, 0xa1, 0x88, 0x8e, 0x2b,
64 0x94, 0x99, 0xb7, 0x71, 0x74, 0xd3, 0xe4, 0xbf,
65 0x3a, 0xde, 0x96, 0x0e, 0xbc, 0x0a, 0xed, 0x77,
66 0xfc, 0x37, 0x6b, 0x03, 0x79, 0x89, 0x62, 0xc6,
67 0xd7, 0xc0, 0xd2, 0x7c, 0x6a, 0x8b, 0x22, 0xa3,
68 0x5b, 0x05, 0x5d, 0x02, 0x75, 0xd5, 0x61, 0xe3,
69 0x18, 0x8f, 0x55, 0x51, 0xad, 0x1f, 0x0b, 0x5e,
70 0x85, 0xe5, 0xc2, 0x57, 0x63, 0xca, 0x3d, 0x6c,
71 0xb4, 0xc5, 0xcc, 0x70, 0xb2, 0x91, 0x59, 0x0d,
72 0x47, 0x20, 0xc8, 0x4f, 0x58, 0xe0, 0x01, 0xe2,
73 0x16, 0x38, 0xc4, 0x6f, 0x3b, 0x0f, 0x65, 0x46,
74 0xbe, 0x7e, 0x2d, 0x7b, 0x82, 0xf9, 0x40, 0xb5,
75 0x1d, 0x73, 0xf8, 0xeb, 0x26, 0xc7, 0x87, 0x97,
76 0x25, 0x54, 0xb1, 0x28, 0xaa, 0x98, 0x9d, 0xa5,
77 0x64, 0x6d, 0x7a, 0xd4, 0x10, 0x81, 0x44, 0xef,
78 0x49, 0xd6, 0xae, 0x2e, 0xdd, 0x76, 0x5c, 0x2f,
79 0xa7, 0x1c, 0xc9, 0x09, 0x69, 0x9a, 0x83, 0xcf,
80 0x29, 0x39, 0xb9, 0xe9, 0x4c, 0xff, 0x43, 0xab,
84 #define BOX(S,i) (u32x) ((S)[(i)])
88 #define BOX(S,i) (u32x) ((S)[(i).s0], (S)[(i).s1])
92 #define BOX(S,i) (u32x) ((S)[(i).s0], (S)[(i).s1], (S)[(i).s2], (S)[(i).s3])
96 #define uint_to_hex_upper8(i) (u32x) (l_bin2asc[(i)])
100 #define uint_to_hex_upper8(i) (u32x) (l_bin2asc[(i).s0], l_bin2asc[(i).s1])
104 #define uint_to_hex_upper8(i) (u32x) (l_bin2asc[(i).s0], l_bin2asc[(i).s1], l_bin2asc[(i).s2], l_bin2asc[(i).s3])
107 static void lotus_mix (u32x *in, __local u32 s_lotus_magic_table[256])
111 for (int i = 0; i < 18; i++)
116 for (int j = 0; j < 12; j++)
121 p = (p + s--) & 0xff; p = ((tmp_in >> 0) & 0xff) ^ BOX (s_lotus_magic_table, p); tmp_out |= p << 0;
122 p = (p + s--) & 0xff; p = ((tmp_in >> 8) & 0xff) ^ BOX (s_lotus_magic_table, p); tmp_out |= p << 8;
123 p = (p + s--) & 0xff; p = ((tmp_in >> 16) & 0xff) ^ BOX (s_lotus_magic_table, p); tmp_out |= p << 16;
124 p = (p + s--) & 0xff; p = ((tmp_in >> 24) & 0xff) ^ BOX (s_lotus_magic_table, p); tmp_out |= p << 24;
131 static void lotus_transform_password (u32x *in, u32x *out, __local u32 s_lotus_magic_table[256])
133 u32x t = out[3] >> 24;
137 //#pragma unroll // kernel fails if used
138 for (int i = 0; i < 4; i++)
140 t ^= (in[i] >> 0) & 0xff; c = BOX (s_lotus_magic_table, t); out[i] ^= c << 0; t = ((out[i] >> 0) & 0xff);
141 t ^= (in[i] >> 8) & 0xff; c = BOX (s_lotus_magic_table, t); out[i] ^= c << 8; t = ((out[i] >> 8) & 0xff);
142 t ^= (in[i] >> 16) & 0xff; c = BOX (s_lotus_magic_table, t); out[i] ^= c << 16; t = ((out[i] >> 16) & 0xff);
143 t ^= (in[i] >> 24) & 0xff; c = BOX (s_lotus_magic_table, t); out[i] ^= c << 24; t = ((out[i] >> 24) & 0xff);
147 static void pad (u32x w[4], const u32 len)
149 const u32 val = 16 - len;
151 const u32 mask1 = val << 24;
153 const u32 mask2 = val << 16
156 const u32 mask3 = val << 8
160 const u32 mask4 = val << 0
167 case 0: w[0] = mask4;
172 case 1: w[0] |= mask3;
177 case 2: w[0] |= mask2;
182 case 3: w[0] |= mask1;
187 case 4: w[1] = mask4;
191 case 5: w[1] |= mask3;
195 case 6: w[1] |= mask2;
199 case 7: w[1] |= mask1;
203 case 8: w[2] = mask4;
206 case 9: w[2] |= mask3;
209 case 10: w[2] |= mask2;
212 case 11: w[2] |= mask1;
215 case 12: w[3] = mask4;
217 case 13: w[3] |= mask3;
219 case 14: w[3] |= mask2;
221 case 15: w[3] |= mask1;
226 static void mdtransform_norecalc (u32x state[4], u32x block[4], __local u32 s_lotus_magic_table[256])
238 x[ 8] = state[0] ^ block[0];
239 x[ 9] = state[1] ^ block[1];
240 x[10] = state[2] ^ block[2];
241 x[11] = state[3] ^ block[3];
243 lotus_mix (x, s_lotus_magic_table);
251 static void mdtransform (u32x state[4], u32x checksum[4], u32x block[4], __local u32 s_lotus_magic_table[256])
253 mdtransform_norecalc (state, block, s_lotus_magic_table);
255 lotus_transform_password (block, checksum, s_lotus_magic_table);
258 static void domino_big_md (const u32x saved_key[16], const u32 size, u32x state[4], __local u32 s_lotus_magic_table[256])
277 for (curpos = 0, idx = 0; curpos + 16 < size; curpos += 16, idx += 4)
279 block[0] = saved_key[idx + 0];
280 block[1] = saved_key[idx + 1];
281 block[2] = saved_key[idx + 2];
282 block[3] = saved_key[idx + 3];
284 mdtransform (state, checksum, block, s_lotus_magic_table);
287 block[0] = saved_key[idx + 0];
288 block[1] = saved_key[idx + 1];
289 block[2] = saved_key[idx + 2];
290 block[3] = saved_key[idx + 3];
292 mdtransform (state, checksum, block, s_lotus_magic_table);
294 mdtransform_norecalc (state, checksum, s_lotus_magic_table);
297 __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m08700_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 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 combs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
303 const u32 lid = get_local_id (0);
309 const u32 gid = get_global_id (0);
315 __local u32 l_bin2asc[256];
317 const u32 lid4 = lid * 4;
319 const u32 lid40 = lid4 + 0;
320 const u32 lid41 = lid4 + 1;
321 const u32 lid42 = lid4 + 2;
322 const u32 lid43 = lid4 + 3;
324 const u32 v400 = (lid40 >> 0) & 15;
325 const u32 v401 = (lid40 >> 4) & 15;
326 const u32 v410 = (lid41 >> 0) & 15;
327 const u32 v411 = (lid41 >> 4) & 15;
328 const u32 v420 = (lid42 >> 0) & 15;
329 const u32 v421 = (lid42 >> 4) & 15;
330 const u32 v430 = (lid43 >> 0) & 15;
331 const u32 v431 = (lid43 >> 4) & 15;
333 l_bin2asc[lid40] = ((v400 < 10) ? '0' + v400 : 'A' - 10 + v400) << 8
334 | ((v401 < 10) ? '0' + v401 : 'A' - 10 + v401) << 0;
335 l_bin2asc[lid41] = ((v410 < 10) ? '0' + v410 : 'A' - 10 + v410) << 8
336 | ((v411 < 10) ? '0' + v411 : 'A' - 10 + v411) << 0;
337 l_bin2asc[lid42] = ((v420 < 10) ? '0' + v420 : 'A' - 10 + v420) << 8
338 | ((v421 < 10) ? '0' + v421 : 'A' - 10 + v421) << 0;
339 l_bin2asc[lid43] = ((v430 < 10) ? '0' + v430 : 'A' - 10 + v430) << 8
340 | ((v431 < 10) ? '0' + v431 : 'A' - 10 + v431) << 0;
342 barrier (CLK_LOCAL_MEM_FENCE);
348 __local u32 s_lotus_magic_table[256];
350 s_lotus_magic_table[lid4 + 0] = lotus_magic_table[lid4 + 0];
351 s_lotus_magic_table[lid4 + 1] = lotus_magic_table[lid4 + 1];
352 s_lotus_magic_table[lid4 + 2] = lotus_magic_table[lid4 + 2];
353 s_lotus_magic_table[lid4 + 3] = lotus_magic_table[lid4 + 3];
355 barrier (CLK_LOCAL_MEM_FENCE);
357 if (gid >= gid_max) return;
361 wordl0[0] = pws[gid].i[ 0];
362 wordl0[1] = pws[gid].i[ 1];
363 wordl0[2] = pws[gid].i[ 2];
364 wordl0[3] = pws[gid].i[ 3];
368 wordl1[0] = pws[gid].i[ 4];
369 wordl1[1] = pws[gid].i[ 5];
370 wordl1[2] = pws[gid].i[ 6];
371 wordl1[3] = pws[gid].i[ 7];
388 const u32 pw_l_len = pws[gid].pw_len;
390 if (combs_mode == COMBINATOR_MODE_BASE_RIGHT)
392 switch_buffer_by_offset (wordl0, wordl1, wordl2, wordl3, combs_buf[0].pw_len);
399 const u32 salt0 = salt_bufs[salt_pos].salt_buf[0];
400 const u32 salt1 = salt_bufs[salt_pos].salt_buf[1] & 0xff | '(' << 8;
406 for (u32 il_pos = 0; il_pos < combs_cnt; il_pos++)
408 const u32 pw_r_len = combs_buf[il_pos].pw_len;
410 const u32 pw_len = pw_l_len + pw_r_len;
414 wordr0[0] = combs_buf[il_pos].i[0];
415 wordr0[1] = combs_buf[il_pos].i[1];
416 wordr0[2] = combs_buf[il_pos].i[2];
417 wordr0[3] = combs_buf[il_pos].i[3];
421 wordr1[0] = combs_buf[il_pos].i[4];
422 wordr1[1] = combs_buf[il_pos].i[5];
423 wordr1[2] = combs_buf[il_pos].i[6];
424 wordr1[3] = combs_buf[il_pos].i[7];
440 if (combs_mode == COMBINATOR_MODE_BASE_LEFT)
442 switch_buffer_by_offset (wordr0, wordr1, wordr2, wordr3, pw_l_len);
447 w[ 0] = wordl0[0] | wordr0[0];
448 w[ 1] = wordl0[1] | wordr0[1];
449 w[ 2] = wordl0[2] | wordr0[2];
450 w[ 3] = wordl0[3] | wordr0[3];
451 w[ 4] = wordl1[0] | wordr1[0];
452 w[ 5] = wordl1[1] | wordr1[1];
453 w[ 6] = wordl1[2] | wordr1[2];
454 w[ 7] = wordl1[3] | wordr1[3];
455 w[ 8] = wordl2[0] | wordr2[0];
456 w[ 9] = wordl2[1] | wordr2[1];
457 w[10] = wordl2[2] | wordr2[2];
458 w[11] = wordl2[3] | wordr2[3];
459 w[12] = wordl3[0] | wordr3[0];
460 w[13] = wordl3[1] | wordr3[1];
461 w[14] = wordl3[2] | wordr3[2];
462 w[15] = wordl3[3] | wordr3[3];
477 pad (&w[ 0], pw_len & 0xf);
479 else if (pw_len < 32)
481 pad (&w[ 4], pw_len & 0xf);
483 else if (pw_len < 48)
485 pad (&w[ 8], pw_len & 0xf);
487 else if (pw_len < 64)
489 pad (&w[12], pw_len & 0xf);
492 domino_big_md (w, pw_len, state, s_lotus_magic_table);
494 const u32x w0_t = uint_to_hex_upper8 ((state[0] >> 0) & 255) << 0
495 | uint_to_hex_upper8 ((state[0] >> 8) & 255) << 16;
496 const u32x w1_t = uint_to_hex_upper8 ((state[0] >> 16) & 255) << 0
497 | uint_to_hex_upper8 ((state[0] >> 24) & 255) << 16;
498 const u32x w2_t = uint_to_hex_upper8 ((state[1] >> 0) & 255) << 0
499 | uint_to_hex_upper8 ((state[1] >> 8) & 255) << 16;
500 const u32x w3_t = uint_to_hex_upper8 ((state[1] >> 16) & 255) << 0
501 | uint_to_hex_upper8 ((state[1] >> 24) & 255) << 16;
502 const u32x w4_t = uint_to_hex_upper8 ((state[2] >> 0) & 255) << 0
503 | uint_to_hex_upper8 ((state[2] >> 8) & 255) << 16;
504 const u32x w5_t = uint_to_hex_upper8 ((state[2] >> 16) & 255) << 0
505 | uint_to_hex_upper8 ((state[2] >> 24) & 255) << 16;
506 const u32x w6_t = uint_to_hex_upper8 ((state[3] >> 0) & 255) << 0
507 | uint_to_hex_upper8 ((state[3] >> 8) & 255) << 16;
508 //const u32x w7_t = uint_to_hex_upper8 ((state[3] >> 16) & 255) << 0
509 // | uint_to_hex_upper8 ((state[3] >> 24) & 255) << 16;
511 const u32 pade = 0x0e0e0e0e;
514 w[ 1] = salt1 | w0_t << 16;
515 w[ 2] = w0_t >> 16 | w1_t << 16;
516 w[ 3] = w1_t >> 16 | w2_t << 16;
517 w[ 4] = w2_t >> 16 | w3_t << 16;
518 w[ 5] = w3_t >> 16 | w4_t << 16;
519 w[ 6] = w4_t >> 16 | w5_t << 16;
520 w[ 7] = w5_t >> 16 | w6_t << 16;
521 w[ 8] = w6_t >> 16 | pade << 16; // | w7_t << 8;
535 domino_big_md (w, 34, state, s_lotus_magic_table);
537 u32x a = state[0] & 0xffffffff;
538 u32x b = state[1] & 0xffffffff;
539 u32x c = state[2] & 0x000000ff;
540 u32x d = state[3] & 0x00000000;
547 #include VECT_COMPARE_M
551 __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m08700_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 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 combs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
555 __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m08700_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 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 combs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
559 __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m08700_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 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 combs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
565 const u32 lid = get_local_id (0);
571 const u32 gid = get_global_id (0);
577 __local u32 l_bin2asc[256];
579 const u32 lid4 = lid * 4;
581 const u32 lid40 = lid4 + 0;
582 const u32 lid41 = lid4 + 1;
583 const u32 lid42 = lid4 + 2;
584 const u32 lid43 = lid4 + 3;
586 const u32 v400 = (lid40 >> 0) & 15;
587 const u32 v401 = (lid40 >> 4) & 15;
588 const u32 v410 = (lid41 >> 0) & 15;
589 const u32 v411 = (lid41 >> 4) & 15;
590 const u32 v420 = (lid42 >> 0) & 15;
591 const u32 v421 = (lid42 >> 4) & 15;
592 const u32 v430 = (lid43 >> 0) & 15;
593 const u32 v431 = (lid43 >> 4) & 15;
595 l_bin2asc[lid40] = ((v400 < 10) ? '0' + v400 : 'A' - 10 + v400) << 8
596 | ((v401 < 10) ? '0' + v401 : 'A' - 10 + v401) << 0;
597 l_bin2asc[lid41] = ((v410 < 10) ? '0' + v410 : 'A' - 10 + v410) << 8
598 | ((v411 < 10) ? '0' + v411 : 'A' - 10 + v411) << 0;
599 l_bin2asc[lid42] = ((v420 < 10) ? '0' + v420 : 'A' - 10 + v420) << 8
600 | ((v421 < 10) ? '0' + v421 : 'A' - 10 + v421) << 0;
601 l_bin2asc[lid43] = ((v430 < 10) ? '0' + v430 : 'A' - 10 + v430) << 8
602 | ((v431 < 10) ? '0' + v431 : 'A' - 10 + v431) << 0;
604 barrier (CLK_LOCAL_MEM_FENCE);
610 __local u32 s_lotus_magic_table[256];
612 s_lotus_magic_table[lid4 + 0] = lotus_magic_table[lid4 + 0];
613 s_lotus_magic_table[lid4 + 1] = lotus_magic_table[lid4 + 1];
614 s_lotus_magic_table[lid4 + 2] = lotus_magic_table[lid4 + 2];
615 s_lotus_magic_table[lid4 + 3] = lotus_magic_table[lid4 + 3];
617 barrier (CLK_LOCAL_MEM_FENCE);
619 if (gid >= gid_max) return;
623 wordl0[0] = pws[gid].i[ 0];
624 wordl0[1] = pws[gid].i[ 1];
625 wordl0[2] = pws[gid].i[ 2];
626 wordl0[3] = pws[gid].i[ 3];
630 wordl1[0] = pws[gid].i[ 4];
631 wordl1[1] = pws[gid].i[ 5];
632 wordl1[2] = pws[gid].i[ 6];
633 wordl1[3] = pws[gid].i[ 7];
650 const u32 pw_l_len = pws[gid].pw_len;
652 if (combs_mode == COMBINATOR_MODE_BASE_RIGHT)
654 switch_buffer_by_offset (wordl0, wordl1, wordl2, wordl3, combs_buf[0].pw_len);
661 const u32 salt0 = salt_bufs[salt_pos].salt_buf[0];
662 const u32 salt1 = salt_bufs[salt_pos].salt_buf[1] & 0xff | '(' << 8;
668 const u32 search[4] =
670 digests_buf[digests_offset].digest_buf[DGST_R0],
671 digests_buf[digests_offset].digest_buf[DGST_R1],
672 digests_buf[digests_offset].digest_buf[DGST_R2],
673 digests_buf[digests_offset].digest_buf[DGST_R3]
680 for (u32 il_pos = 0; il_pos < combs_cnt; il_pos++)
682 const u32 pw_r_len = combs_buf[il_pos].pw_len;
684 const u32 pw_len = pw_l_len + pw_r_len;
688 wordr0[0] = combs_buf[il_pos].i[0];
689 wordr0[1] = combs_buf[il_pos].i[1];
690 wordr0[2] = combs_buf[il_pos].i[2];
691 wordr0[3] = combs_buf[il_pos].i[3];
695 wordr1[0] = combs_buf[il_pos].i[4];
696 wordr1[1] = combs_buf[il_pos].i[5];
697 wordr1[2] = combs_buf[il_pos].i[6];
698 wordr1[3] = combs_buf[il_pos].i[7];
714 if (combs_mode == COMBINATOR_MODE_BASE_LEFT)
716 switch_buffer_by_offset (wordr0, wordr1, wordr2, wordr3, pw_l_len);
721 w[ 0] = wordl0[0] | wordr0[0];
722 w[ 1] = wordl0[1] | wordr0[1];
723 w[ 2] = wordl0[2] | wordr0[2];
724 w[ 3] = wordl0[3] | wordr0[3];
725 w[ 4] = wordl1[0] | wordr1[0];
726 w[ 5] = wordl1[1] | wordr1[1];
727 w[ 6] = wordl1[2] | wordr1[2];
728 w[ 7] = wordl1[3] | wordr1[3];
729 w[ 8] = wordl2[0] | wordr2[0];
730 w[ 9] = wordl2[1] | wordr2[1];
731 w[10] = wordl2[2] | wordr2[2];
732 w[11] = wordl2[3] | wordr2[3];
733 w[12] = wordl3[0] | wordr3[0];
734 w[13] = wordl3[1] | wordr3[1];
735 w[14] = wordl3[2] | wordr3[2];
736 w[15] = wordl3[3] | wordr3[3];
751 pad (&w[ 0], pw_len & 0xf);
753 else if (pw_len < 32)
755 pad (&w[ 4], pw_len & 0xf);
757 else if (pw_len < 48)
759 pad (&w[ 8], pw_len & 0xf);
761 else if (pw_len < 64)
763 pad (&w[12], pw_len & 0xf);
766 domino_big_md (w, pw_len, state, s_lotus_magic_table);
768 const u32x w0_t = uint_to_hex_upper8 ((state[0] >> 0) & 255) << 0
769 | uint_to_hex_upper8 ((state[0] >> 8) & 255) << 16;
770 const u32x w1_t = uint_to_hex_upper8 ((state[0] >> 16) & 255) << 0
771 | uint_to_hex_upper8 ((state[0] >> 24) & 255) << 16;
772 const u32x w2_t = uint_to_hex_upper8 ((state[1] >> 0) & 255) << 0
773 | uint_to_hex_upper8 ((state[1] >> 8) & 255) << 16;
774 const u32x w3_t = uint_to_hex_upper8 ((state[1] >> 16) & 255) << 0
775 | uint_to_hex_upper8 ((state[1] >> 24) & 255) << 16;
776 const u32x w4_t = uint_to_hex_upper8 ((state[2] >> 0) & 255) << 0
777 | uint_to_hex_upper8 ((state[2] >> 8) & 255) << 16;
778 const u32x w5_t = uint_to_hex_upper8 ((state[2] >> 16) & 255) << 0
779 | uint_to_hex_upper8 ((state[2] >> 24) & 255) << 16;
780 const u32x w6_t = uint_to_hex_upper8 ((state[3] >> 0) & 255) << 0
781 | uint_to_hex_upper8 ((state[3] >> 8) & 255) << 16;
782 //const u32x w7_t = uint_to_hex_upper8 ((state[3] >> 16) & 255) << 0
783 // | uint_to_hex_upper8 ((state[3] >> 24) & 255) << 16;
785 const u32 pade = 0x0e0e0e0e;
788 w[ 1] = salt1 | w0_t << 16;
789 w[ 2] = w0_t >> 16 | w1_t << 16;
790 w[ 3] = w1_t >> 16 | w2_t << 16;
791 w[ 4] = w2_t >> 16 | w3_t << 16;
792 w[ 5] = w3_t >> 16 | w4_t << 16;
793 w[ 6] = w4_t >> 16 | w5_t << 16;
794 w[ 7] = w5_t >> 16 | w6_t << 16;
795 w[ 8] = w6_t >> 16 | pade << 16; // | w7_t << 8;
809 domino_big_md (w, 34, state, s_lotus_magic_table);
811 u32x a = state[0] & 0xffffffff;
812 u32x b = state[1] & 0xffffffff;
813 u32x c = state[2] & 0x000000ff;
814 u32x d = state[3] & 0x00000000;
821 #include VECT_COMPARE_S
825 __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m08700_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 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 combs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
829 __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m08700_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 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 combs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)