2 * Author......: Jens Steube <jens.steube@gmail.com>
8 //incompatible because of branches
9 //#define NEW_SIMD_CODE
11 #include "inc_vendor.cl"
12 #include "inc_hash_constants.h"
13 #include "inc_hash_functions.cl"
14 #include "inc_types.cl"
15 #include "inc_common.cl"
16 #include "inc_simd.cl"
18 __constant u32 crc32tab[0x100] =
20 0x00000000, 0x77073096, 0xee0e612c, 0x990951ba,
21 0x076dc419, 0x706af48f, 0xe963a535, 0x9e6495a3,
22 0x0edb8832, 0x79dcb8a4, 0xe0d5e91e, 0x97d2d988,
23 0x09b64c2b, 0x7eb17cbd, 0xe7b82d07, 0x90bf1d91,
24 0x1db71064, 0x6ab020f2, 0xf3b97148, 0x84be41de,
25 0x1adad47d, 0x6ddde4eb, 0xf4d4b551, 0x83d385c7,
26 0x136c9856, 0x646ba8c0, 0xfd62f97a, 0x8a65c9ec,
27 0x14015c4f, 0x63066cd9, 0xfa0f3d63, 0x8d080df5,
28 0x3b6e20c8, 0x4c69105e, 0xd56041e4, 0xa2677172,
29 0x3c03e4d1, 0x4b04d447, 0xd20d85fd, 0xa50ab56b,
30 0x35b5a8fa, 0x42b2986c, 0xdbbbc9d6, 0xacbcf940,
31 0x32d86ce3, 0x45df5c75, 0xdcd60dcf, 0xabd13d59,
32 0x26d930ac, 0x51de003a, 0xc8d75180, 0xbfd06116,
33 0x21b4f4b5, 0x56b3c423, 0xcfba9599, 0xb8bda50f,
34 0x2802b89e, 0x5f058808, 0xc60cd9b2, 0xb10be924,
35 0x2f6f7c87, 0x58684c11, 0xc1611dab, 0xb6662d3d,
36 0x76dc4190, 0x01db7106, 0x98d220bc, 0xefd5102a,
37 0x71b18589, 0x06b6b51f, 0x9fbfe4a5, 0xe8b8d433,
38 0x7807c9a2, 0x0f00f934, 0x9609a88e, 0xe10e9818,
39 0x7f6a0dbb, 0x086d3d2d, 0x91646c97, 0xe6635c01,
40 0x6b6b51f4, 0x1c6c6162, 0x856530d8, 0xf262004e,
41 0x6c0695ed, 0x1b01a57b, 0x8208f4c1, 0xf50fc457,
42 0x65b0d9c6, 0x12b7e950, 0x8bbeb8ea, 0xfcb9887c,
43 0x62dd1ddf, 0x15da2d49, 0x8cd37cf3, 0xfbd44c65,
44 0x4db26158, 0x3ab551ce, 0xa3bc0074, 0xd4bb30e2,
45 0x4adfa541, 0x3dd895d7, 0xa4d1c46d, 0xd3d6f4fb,
46 0x4369e96a, 0x346ed9fc, 0xad678846, 0xda60b8d0,
47 0x44042d73, 0x33031de5, 0xaa0a4c5f, 0xdd0d7cc9,
48 0x5005713c, 0x270241aa, 0xbe0b1010, 0xc90c2086,
49 0x5768b525, 0x206f85b3, 0xb966d409, 0xce61e49f,
50 0x5edef90e, 0x29d9c998, 0xb0d09822, 0xc7d7a8b4,
51 0x59b33d17, 0x2eb40d81, 0xb7bd5c3b, 0xc0ba6cad,
52 0xedb88320, 0x9abfb3b6, 0x03b6e20c, 0x74b1d29a,
53 0xead54739, 0x9dd277af, 0x04db2615, 0x73dc1683,
54 0xe3630b12, 0x94643b84, 0x0d6d6a3e, 0x7a6a5aa8,
55 0xe40ecf0b, 0x9309ff9d, 0x0a00ae27, 0x7d079eb1,
56 0xf00f9344, 0x8708a3d2, 0x1e01f268, 0x6906c2fe,
57 0xf762575d, 0x806567cb, 0x196c3671, 0x6e6b06e7,
58 0xfed41b76, 0x89d32be0, 0x10da7a5a, 0x67dd4acc,
59 0xf9b9df6f, 0x8ebeeff9, 0x17b7be43, 0x60b08ed5,
60 0xd6d6a3e8, 0xa1d1937e, 0x38d8c2c4, 0x4fdff252,
61 0xd1bb67f1, 0xa6bc5767, 0x3fb506dd, 0x48b2364b,
62 0xd80d2bda, 0xaf0a1b4c, 0x36034af6, 0x41047a60,
63 0xdf60efc3, 0xa867df55, 0x316e8eef, 0x4669be79,
64 0xcb61b38c, 0xbc66831a, 0x256fd2a0, 0x5268e236,
65 0xcc0c7795, 0xbb0b4703, 0x220216b9, 0x5505262f,
66 0xc5ba3bbe, 0xb2bd0b28, 0x2bb45a92, 0x5cb36a04,
67 0xc2d7ffa7, 0xb5d0cf31, 0x2cd99e8b, 0x5bdeae1d,
68 0x9b64c2b0, 0xec63f226, 0x756aa39c, 0x026d930a,
69 0x9c0906a9, 0xeb0e363f, 0x72076785, 0x05005713,
70 0x95bf4a82, 0xe2b87a14, 0x7bb12bae, 0x0cb61b38,
71 0x92d28e9b, 0xe5d5be0d, 0x7cdcefb7, 0x0bdbdf21,
72 0x86d3d2d4, 0xf1d4e242, 0x68ddb3f8, 0x1fda836e,
73 0x81be16cd, 0xf6b9265b, 0x6fb077e1, 0x18b74777,
74 0x88085ae6, 0xff0f6a70, 0x66063bca, 0x11010b5c,
75 0x8f659eff, 0xf862ae69, 0x616bffd3, 0x166ccf45,
76 0xa00ae278, 0xd70dd2ee, 0x4e048354, 0x3903b3c2,
77 0xa7672661, 0xd06016f7, 0x4969474d, 0x3e6e77db,
78 0xaed16a4a, 0xd9d65adc, 0x40df0b66, 0x37d83bf0,
79 0xa9bcae53, 0xdebb9ec5, 0x47b2cf7f, 0x30b5ffe9,
80 0xbdbdf21c, 0xcabac28a, 0x53b39330, 0x24b4a3a6,
81 0xbad03605, 0xcdd70693, 0x54de5729, 0x23d967bf,
82 0xb3667a2e, 0xc4614ab8, 0x5d681b02, 0x2a6f2b94,
83 0xb40bbe37, 0xc30c8ea1, 0x5a05df1b, 0x2d02ef8d
86 u32x round_crc32 (u32x a, const u32x v)
88 const u32x k = (a ^ v) & 0xff;
90 const u32x s = a >> 8;
93 a = (u32x) crc32tab[k];
95 a = (u32x) (crc32tab[k.s0], crc32tab[k.s1]);
97 a = (u32x) (crc32tab[k.s0], crc32tab[k.s1], crc32tab[k.s2], crc32tab[k.s3]);
99 a = (u32x) (crc32tab[k.s0], crc32tab[k.s1], crc32tab[k.s2], crc32tab[k.s3], crc32tab[k.s4], crc32tab[k.s5], crc32tab[k.s6], crc32tab[k.s7]);
100 #elif VECT_SIZE == 16
101 a = (u32x) (crc32tab[k.s0], crc32tab[k.s1], crc32tab[k.s2], crc32tab[k.s3], crc32tab[k.s4], crc32tab[k.s5], crc32tab[k.s6], crc32tab[k.s7], crc32tab[k.s8], crc32tab[k.s9], crc32tab[k.sa], crc32tab[k.sb], crc32tab[k.sc], crc32tab[k.sd], crc32tab[k.se], crc32tab[k.sf]);
109 u32x crc32 (const u32x w[16], const u32 pw_len, const u32 iv)
113 if (pw_len >= 1) a = round_crc32 (a, w[0] >> 0);
114 if (pw_len >= 2) a = round_crc32 (a, w[0] >> 8);
115 if (pw_len >= 3) a = round_crc32 (a, w[0] >> 16);
116 if (pw_len >= 4) a = round_crc32 (a, w[0] >> 24);
118 for (u32 i = 4, j = 1; i < pw_len; i += 4, j += 1)
120 if (pw_len >= (i + 1)) a = round_crc32 (a, w[j] >> 0);
121 if (pw_len >= (i + 2)) a = round_crc32 (a, w[j] >> 8);
122 if (pw_len >= (i + 3)) a = round_crc32 (a, w[j] >> 16);
123 if (pw_len >= (i + 4)) a = round_crc32 (a, w[j] >> 24);
129 __kernel void m11500_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 void *esalt_bufs, __global u32 *d_return_buf, __global u32 *d_scryptV0_buf, __global u32 *d_scryptV1_buf, __global u32 *d_scryptV2_buf, __global u32 *d_scryptV3_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)
135 const u32 lid = get_local_id (0);
141 const u32 gid = get_global_id (0);
143 if (gid >= gid_max) return;
148 pw_buf0[0] = pws[gid].i[0];
149 pw_buf0[1] = pws[gid].i[1];
150 pw_buf0[2] = pws[gid].i[2];
151 pw_buf0[3] = pws[gid].i[3];
152 pw_buf1[0] = pws[gid].i[4];
153 pw_buf1[1] = pws[gid].i[5];
154 pw_buf1[2] = pws[gid].i[6];
155 pw_buf1[3] = pws[gid].i[7];
157 const u32 pw_l_len = pws[gid].pw_len;
163 const u32 iv = salt_bufs[salt_pos].salt_buf[0];
169 for (u32 il_pos = 0; il_pos < il_cnt; il_pos += VECT_SIZE)
171 const u32x pw_r_len = pwlenx_create_combt (combs_buf, il_pos);
173 const u32x pw_len = pw_l_len + pw_r_len;
176 * concat password candidate
179 u32x wordl0[4] = { 0 };
180 u32x wordl1[4] = { 0 };
181 u32x wordl2[4] = { 0 };
182 u32x wordl3[4] = { 0 };
184 wordl0[0] = pw_buf0[0];
185 wordl0[1] = pw_buf0[1];
186 wordl0[2] = pw_buf0[2];
187 wordl0[3] = pw_buf0[3];
188 wordl1[0] = pw_buf1[0];
189 wordl1[1] = pw_buf1[1];
190 wordl1[2] = pw_buf1[2];
191 wordl1[3] = pw_buf1[3];
193 u32x wordr0[4] = { 0 };
194 u32x wordr1[4] = { 0 };
195 u32x wordr2[4] = { 0 };
196 u32x wordr3[4] = { 0 };
198 wordr0[0] = ix_create_combt (combs_buf, il_pos, 0);
199 wordr0[1] = ix_create_combt (combs_buf, il_pos, 1);
200 wordr0[2] = ix_create_combt (combs_buf, il_pos, 2);
201 wordr0[3] = ix_create_combt (combs_buf, il_pos, 3);
202 wordr1[0] = ix_create_combt (combs_buf, il_pos, 4);
203 wordr1[1] = ix_create_combt (combs_buf, il_pos, 5);
204 wordr1[2] = ix_create_combt (combs_buf, il_pos, 6);
205 wordr1[3] = ix_create_combt (combs_buf, il_pos, 7);
207 if (combs_mode == COMBINATOR_MODE_BASE_LEFT)
209 switch_buffer_by_offset_le_VV (wordr0, wordr1, wordr2, wordr3, pw_l_len);
213 switch_buffer_by_offset_le_VV (wordl0, wordl1, wordl2, wordl3, pw_r_len);
221 w0[0] = wordl0[0] | wordr0[0];
222 w0[1] = wordl0[1] | wordr0[1];
223 w0[2] = wordl0[2] | wordr0[2];
224 w0[3] = wordl0[3] | wordr0[3];
225 w1[0] = wordl1[0] | wordr1[0];
226 w1[1] = wordl1[1] | wordr1[1];
227 w1[2] = wordl1[2] | wordr1[2];
228 w1[3] = wordl1[3] | wordr1[3];
229 w2[0] = wordl2[0] | wordr2[0];
230 w2[1] = wordl2[1] | wordr2[1];
231 w2[2] = wordl2[2] | wordr2[2];
232 w2[3] = wordl2[3] | wordr2[3];
233 w3[0] = wordl3[0] | wordr3[0];
234 w3[1] = wordl3[1] | wordr3[1];
235 w3[2] = wordl3[2] | wordr3[2];
236 w3[3] = wordl3[3] | wordr3[3];
261 u32x a = crc32 (w, pw_len, iv);
265 COMPARE_M_SIMD (a, z, z, z);
269 __kernel void m11500_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 void *esalt_bufs, __global u32 *d_return_buf, __global u32 *d_scryptV0_buf, __global u32 *d_scryptV1_buf, __global u32 *d_scryptV2_buf, __global u32 *d_scryptV3_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)
273 __kernel void m11500_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 void *esalt_bufs, __global u32 *d_return_buf, __global u32 *d_scryptV0_buf, __global u32 *d_scryptV1_buf, __global u32 *d_scryptV2_buf, __global u32 *d_scryptV3_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)
277 __kernel void m11500_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 void *esalt_bufs, __global u32 *d_return_buf, __global u32 *d_scryptV0_buf, __global u32 *d_scryptV1_buf, __global u32 *d_scryptV2_buf, __global u32 *d_scryptV3_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)
283 const u32 lid = get_local_id (0);
289 const u32 gid = get_global_id (0);
291 if (gid >= gid_max) return;
296 pw_buf0[0] = pws[gid].i[0];
297 pw_buf0[1] = pws[gid].i[1];
298 pw_buf0[2] = pws[gid].i[2];
299 pw_buf0[3] = pws[gid].i[3];
300 pw_buf1[0] = pws[gid].i[4];
301 pw_buf1[1] = pws[gid].i[5];
302 pw_buf1[2] = pws[gid].i[6];
303 pw_buf1[3] = pws[gid].i[7];
305 const u32 pw_l_len = pws[gid].pw_len;
311 const u32 iv = salt_bufs[salt_pos].salt_buf[0];
317 const u32 search[4] =
319 digests_buf[digests_offset].digest_buf[DGST_R0],
329 for (u32 il_pos = 0; il_pos < il_cnt; il_pos += VECT_SIZE)
331 const u32x pw_r_len = pwlenx_create_combt (combs_buf, il_pos);
333 const u32x pw_len = pw_l_len + pw_r_len;
336 * concat password candidate
339 u32x wordl0[4] = { 0 };
340 u32x wordl1[4] = { 0 };
341 u32x wordl2[4] = { 0 };
342 u32x wordl3[4] = { 0 };
344 wordl0[0] = pw_buf0[0];
345 wordl0[1] = pw_buf0[1];
346 wordl0[2] = pw_buf0[2];
347 wordl0[3] = pw_buf0[3];
348 wordl1[0] = pw_buf1[0];
349 wordl1[1] = pw_buf1[1];
350 wordl1[2] = pw_buf1[2];
351 wordl1[3] = pw_buf1[3];
353 u32x wordr0[4] = { 0 };
354 u32x wordr1[4] = { 0 };
355 u32x wordr2[4] = { 0 };
356 u32x wordr3[4] = { 0 };
358 wordr0[0] = ix_create_combt (combs_buf, il_pos, 0);
359 wordr0[1] = ix_create_combt (combs_buf, il_pos, 1);
360 wordr0[2] = ix_create_combt (combs_buf, il_pos, 2);
361 wordr0[3] = ix_create_combt (combs_buf, il_pos, 3);
362 wordr1[0] = ix_create_combt (combs_buf, il_pos, 4);
363 wordr1[1] = ix_create_combt (combs_buf, il_pos, 5);
364 wordr1[2] = ix_create_combt (combs_buf, il_pos, 6);
365 wordr1[3] = ix_create_combt (combs_buf, il_pos, 7);
367 if (combs_mode == COMBINATOR_MODE_BASE_LEFT)
369 switch_buffer_by_offset_le_VV (wordr0, wordr1, wordr2, wordr3, pw_l_len);
373 switch_buffer_by_offset_le_VV (wordl0, wordl1, wordl2, wordl3, pw_r_len);
381 w0[0] = wordl0[0] | wordr0[0];
382 w0[1] = wordl0[1] | wordr0[1];
383 w0[2] = wordl0[2] | wordr0[2];
384 w0[3] = wordl0[3] | wordr0[3];
385 w1[0] = wordl1[0] | wordr1[0];
386 w1[1] = wordl1[1] | wordr1[1];
387 w1[2] = wordl1[2] | wordr1[2];
388 w1[3] = wordl1[3] | wordr1[3];
389 w2[0] = wordl2[0] | wordr2[0];
390 w2[1] = wordl2[1] | wordr2[1];
391 w2[2] = wordl2[2] | wordr2[2];
392 w2[3] = wordl2[3] | wordr2[3];
393 w3[0] = wordl3[0] | wordr3[0];
394 w3[1] = wordl3[1] | wordr3[1];
395 w3[2] = wordl3[2] | wordr3[2];
396 w3[3] = wordl3[3] | wordr3[3];
421 u32x a = crc32 (w, pw_len, iv);
425 COMPARE_S_SIMD (a, z, z, z);
429 __kernel void m11500_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 void *esalt_bufs, __global u32 *d_return_buf, __global u32 *d_scryptV0_buf, __global u32 *d_scryptV1_buf, __global u32 *d_scryptV2_buf, __global u32 *d_scryptV3_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)
433 __kernel void m11500_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 void *esalt_bufs, __global u32 *d_return_buf, __global u32 *d_scryptV0_buf, __global u32 *d_scryptV1_buf, __global u32 *d_scryptV2_buf, __global u32 *d_scryptV3_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)