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"
37 #define GETSHIFTEDINT(a,n) amd_bytealign ((a)[((n)/4)+1], (a)[((n)/4)+0], (n))
39 #define SETSHIFTEDINT(a,n,v) \
41 const u32 s = ((n) & 3) * 8; \
42 const u64 x = (u64) (v) << s; \
43 (a)[((n)/4)+0] |= x; \
44 (a)[((n)/4)+1] = x >> 32; \
47 __constant u32 theMagicArray[64] =
49 0x1451ac91,0x4354679f,0xe03be724,0xc27b7428,0xeb133386,0x5ccb4f5a,0x37730a08,0x2f1c5d0e,
50 0xe5e68f33,0xddae9bf8,0x8d4bf216,0xdcd4e12c,0x9ddfcbb0,0x176d70d4,0x3f424df9,0x94111b9b,
51 0x9bc15b9f,0x039d0506,0x8a135e9d,0xe86a9a1e,0x17147cd9,0xf62ac758,0x0a6399a1,0xc370fdd7,
52 0x13745ef6,0x040bc903,0x26f79826,0x2593928a,0x230da2b0,0x6d7963ed,0x3cfa3213,0xa39a0235,
53 0x0a8eddb3,0xc351bf24,0x9f55cd7c,0x4c94af37,0x82520829,0x374e3bb2,0x9107179f,0xcdfd3b11,
54 0, 0, 0, 0, 0, 0, 0, 0,
55 0, 0, 0, 0, 0, 0, 0, 0,
56 0, 0, 0, 0, 0, 0, 0, 0
59 static void swap_buffer (u32x final[16])
61 final[ 0] = swap_workaround (final[ 0]);
62 final[ 1] = swap_workaround (final[ 1]);
63 final[ 2] = swap_workaround (final[ 2]);
64 final[ 3] = swap_workaround (final[ 3]);
65 final[ 4] = swap_workaround (final[ 4]);
66 final[ 5] = swap_workaround (final[ 5]);
67 final[ 6] = swap_workaround (final[ 6]);
68 final[ 7] = swap_workaround (final[ 7]);
69 final[ 8] = swap_workaround (final[ 8]);
70 final[ 9] = swap_workaround (final[ 9]);
71 final[10] = swap_workaround (final[10]);
72 final[11] = swap_workaround (final[11]);
73 final[12] = swap_workaround (final[12]);
74 final[13] = swap_workaround (final[13]);
75 final[14] = swap_workaround (final[14]);
76 final[15] = swap_workaround (final[15]);
79 static void sha1_transform (const u32x w0[4], const u32x w1[4], const u32x w2[4], const u32x w3[4], u32x digest[5])
107 SHA1_STEP (SHA1_F0o, A, B, C, D, E, w0_t);
108 SHA1_STEP (SHA1_F0o, E, A, B, C, D, w1_t);
109 SHA1_STEP (SHA1_F0o, D, E, A, B, C, w2_t);
110 SHA1_STEP (SHA1_F0o, C, D, E, A, B, w3_t);
111 SHA1_STEP (SHA1_F0o, B, C, D, E, A, w4_t);
112 SHA1_STEP (SHA1_F0o, A, B, C, D, E, w5_t);
113 SHA1_STEP (SHA1_F0o, E, A, B, C, D, w6_t);
114 SHA1_STEP (SHA1_F0o, D, E, A, B, C, w7_t);
115 SHA1_STEP (SHA1_F0o, C, D, E, A, B, w8_t);
116 SHA1_STEP (SHA1_F0o, B, C, D, E, A, w9_t);
117 SHA1_STEP (SHA1_F0o, A, B, C, D, E, wa_t);
118 SHA1_STEP (SHA1_F0o, E, A, B, C, D, wb_t);
119 SHA1_STEP (SHA1_F0o, D, E, A, B, C, wc_t);
120 SHA1_STEP (SHA1_F0o, C, D, E, A, B, wd_t);
121 SHA1_STEP (SHA1_F0o, B, C, D, E, A, we_t);
122 SHA1_STEP (SHA1_F0o, A, B, C, D, E, wf_t);
123 w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F0o, E, A, B, C, D, w0_t);
124 w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F0o, D, E, A, B, C, w1_t);
125 w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F0o, C, D, E, A, B, w2_t);
126 w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F0o, B, C, D, E, A, w3_t);
131 w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w4_t);
132 w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w5_t);
133 w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w6_t);
134 w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w7_t);
135 w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w8_t);
136 w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w9_t);
137 wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wa_t);
138 wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, wb_t);
139 wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, wc_t);
140 wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, wd_t);
141 we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, we_t);
142 wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wf_t);
143 w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w0_t);
144 w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w1_t);
145 w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w2_t);
146 w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w3_t);
147 w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w4_t);
148 w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w5_t);
149 w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w6_t);
150 w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w7_t);
155 w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, w8_t);
156 w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, w9_t);
157 wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, wa_t);
158 wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, wb_t);
159 wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, wc_t);
160 wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, wd_t);
161 we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, we_t);
162 wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, wf_t);
163 w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, w0_t);
164 w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, w1_t);
165 w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, w2_t);
166 w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, w3_t);
167 w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, w4_t);
168 w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, w5_t);
169 w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, w6_t);
170 w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, w7_t);
171 w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, w8_t);
172 w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, w9_t);
173 wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, wa_t);
174 wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, wb_t);
179 wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, wc_t);
180 wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wd_t);
181 we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, we_t);
182 wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, wf_t);
183 w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w0_t);
184 w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w1_t);
185 w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w2_t);
186 w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w3_t);
187 w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w4_t);
188 w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w5_t);
189 w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w6_t);
190 w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w7_t);
191 w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w8_t);
192 w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w9_t);
193 wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, wa_t);
194 wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, wb_t);
195 wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wc_t);
196 wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, wd_t);
197 we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, we_t);
198 wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, wf_t);
207 __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m07800_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)
213 const u32 lid = get_local_id (0);
219 const u32 gid = get_global_id (0);
221 if (gid >= gid_max) return;
225 wordl0[0] = pws[gid].i[ 0];
226 wordl0[1] = pws[gid].i[ 1];
227 wordl0[2] = pws[gid].i[ 2];
228 wordl0[3] = pws[gid].i[ 3];
232 wordl1[0] = pws[gid].i[ 4];
233 wordl1[1] = pws[gid].i[ 5];
234 wordl1[2] = pws[gid].i[ 6];
235 wordl1[3] = pws[gid].i[ 7];
251 const u32 pw_l_len = pws[gid].pw_len;
253 if (combs_mode == COMBINATOR_MODE_BASE_RIGHT)
255 switch_buffer_by_offset (wordl0, wordl1, wordl2, wordl3, combs_buf[0].pw_len);
264 salt_buf[0] = salt_bufs[salt_pos].salt_buf[0];
265 salt_buf[1] = salt_bufs[salt_pos].salt_buf[1];
266 salt_buf[2] = salt_bufs[salt_pos].salt_buf[2];
267 salt_buf[3] = salt_bufs[salt_pos].salt_buf[3];
268 salt_buf[4] = salt_bufs[salt_pos].salt_buf[4];
269 salt_buf[5] = salt_bufs[salt_pos].salt_buf[5];
270 salt_buf[6] = salt_bufs[salt_pos].salt_buf[6];
271 salt_buf[7] = salt_bufs[salt_pos].salt_buf[7];
273 const u32 salt_len = salt_bufs[salt_pos].salt_len;
279 for (u32 il_pos = 0; il_pos < combs_cnt; il_pos++)
281 const u32 pw_r_len = combs_buf[il_pos].pw_len;
283 const u32 pw_len = pw_l_len + pw_r_len;
287 wordr0[0] = combs_buf[il_pos].i[0];
288 wordr0[1] = combs_buf[il_pos].i[1];
289 wordr0[2] = combs_buf[il_pos].i[2];
290 wordr0[3] = combs_buf[il_pos].i[3];
294 wordr1[0] = combs_buf[il_pos].i[4];
295 wordr1[1] = combs_buf[il_pos].i[5];
296 wordr1[2] = combs_buf[il_pos].i[6];
297 wordr1[3] = combs_buf[il_pos].i[7];
313 if (combs_mode == COMBINATOR_MODE_BASE_LEFT)
315 switch_buffer_by_offset (wordr0, wordr1, wordr2, wordr3, pw_l_len);
350 switch_buffer_by_offset (s0, s1, s2, s3, pw_len);
352 const u32 pw_salt_len = pw_len + salt_len;
356 w0[0] = wordl0[0] | wordr0[0];
357 w0[1] = wordl0[1] | wordr0[1];
358 w0[2] = wordl0[2] | wordr0[2];
359 w0[3] = wordl0[3] | wordr0[3];
363 w1[0] = wordl1[0] | wordr1[0];
364 w1[1] = wordl1[1] | wordr1[1];
365 w1[2] = wordl1[2] | wordr1[2];
366 w1[3] = wordl1[3] | wordr1[3];
370 w2[0] = wordl2[0] | wordr2[0];
371 w2[1] = wordl2[1] | wordr2[1];
372 w2[2] = wordl2[2] | wordr2[2];
373 w2[3] = wordl2[3] | wordr2[3];
377 w3[0] = wordl3[0] | wordr3[0];
378 w3[1] = wordl3[1] | wordr3[1];
388 final[ 0] = swap_workaround (w0[0] | s0[0]);
389 final[ 1] = swap_workaround (w0[1] | s0[1]);
390 final[ 2] = swap_workaround (w0[2] | s0[2]);
391 final[ 3] = swap_workaround (w0[3] | s0[3]);
392 final[ 4] = swap_workaround (w1[0] | s1[0]);
393 final[ 5] = swap_workaround (w1[1] | s1[1]);
394 final[ 6] = swap_workaround (w1[2] | s1[2]);
395 final[ 7] = swap_workaround (w1[3] | s1[3]);
396 final[ 8] = swap_workaround (w2[0] | s2[0]);
397 final[ 9] = swap_workaround (w2[1] | s2[1]);
398 final[10] = swap_workaround (w2[2] | s2[2]);
399 final[11] = swap_workaround (w2[3] | s2[3]);
400 final[12] = swap_workaround (w3[0] | s3[0]);
401 final[13] = swap_workaround (w3[1] | s3[1]);
403 final[15] = pw_salt_len * 8;
413 sha1_transform (&final[0], &final[4], &final[8], &final[12], digest);
415 // prepare magic array range
417 u32x lengthMagicArray = 0x20;
418 u32x offsetMagicArray = 0;
420 lengthMagicArray += ((digest[0] >> 24) & 0xff) % 6;
421 lengthMagicArray += ((digest[0] >> 16) & 0xff) % 6;
422 lengthMagicArray += ((digest[0] >> 8) & 0xff) % 6;
423 lengthMagicArray += ((digest[0] >> 0) & 0xff) % 6;
424 lengthMagicArray += ((digest[1] >> 24) & 0xff) % 6;
425 lengthMagicArray += ((digest[1] >> 16) & 0xff) % 6;
426 lengthMagicArray += ((digest[1] >> 8) & 0xff) % 6;
427 lengthMagicArray += ((digest[1] >> 0) & 0xff) % 6;
428 lengthMagicArray += ((digest[2] >> 24) & 0xff) % 6;
429 lengthMagicArray += ((digest[2] >> 16) & 0xff) % 6;
430 offsetMagicArray += ((digest[2] >> 8) & 0xff) % 8;
431 offsetMagicArray += ((digest[2] >> 0) & 0xff) % 8;
432 offsetMagicArray += ((digest[3] >> 24) & 0xff) % 8;
433 offsetMagicArray += ((digest[3] >> 16) & 0xff) % 8;
434 offsetMagicArray += ((digest[3] >> 8) & 0xff) % 8;
435 offsetMagicArray += ((digest[3] >> 0) & 0xff) % 8;
436 offsetMagicArray += ((digest[4] >> 24) & 0xff) % 8;
437 offsetMagicArray += ((digest[4] >> 16) & 0xff) % 8;
438 offsetMagicArray += ((digest[4] >> 8) & 0xff) % 8;
439 offsetMagicArray += ((digest[4] >> 0) & 0xff) % 8;
450 for (int i = 0; i < 64; i++) final[i] = 0;
461 u32 final_len = pw_len;
467 for (i = 0; i < lengthMagicArray - 4; i += 4)
469 const u32 tmp = GETSHIFTEDINT (theMagicArray, offsetMagicArray + i);
471 SETSHIFTEDINT (final, final_len + i, tmp);
474 const u32 mask = 0xffffffff >> (((i - lengthMagicArray) & 3) * 8);
476 const u32 tmp = GETSHIFTEDINT (theMagicArray, offsetMagicArray + i) & mask;
478 SETSHIFTEDINT (final, final_len + i, tmp);
480 final_len += lengthMagicArray;
484 for (i = 0; i < salt_len + 1; i += 4) // +1 for the 0x80
486 const u32 tmp = salt_buf[i / 4]; // attention, int[] not char[]
488 SETSHIFTEDINT (final, final_len + i, tmp);
491 final_len += salt_len;
498 for (left = final_len, off = 0; left >= 56; left -= 64, off += 16)
500 swap_buffer (&final[off]);
502 sha1_transform (&final[off + 0], &final[off + 4], &final[off + 8], &final[off + 12], digest);
505 swap_buffer (&final[off]);
508 final[off + 15] = final_len * 8;
510 sha1_transform (&final[off + 0], &final[off + 4], &final[off + 8], &final[off + 12], digest);
512 const u32x r0 = digest[3];
513 const u32x r1 = digest[4];
514 const u32x r2 = digest[2];
515 const u32x r3 = digest[1];
517 #include VECT_COMPARE_M
521 __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m07800_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)
525 __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m07800_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)
529 __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m07800_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)
535 const u32 lid = get_local_id (0);
541 const u32 gid = get_global_id (0);
543 if (gid >= gid_max) return;
547 wordl0[0] = pws[gid].i[ 0];
548 wordl0[1] = pws[gid].i[ 1];
549 wordl0[2] = pws[gid].i[ 2];
550 wordl0[3] = pws[gid].i[ 3];
554 wordl1[0] = pws[gid].i[ 4];
555 wordl1[1] = pws[gid].i[ 5];
556 wordl1[2] = pws[gid].i[ 6];
557 wordl1[3] = pws[gid].i[ 7];
573 const u32 pw_l_len = pws[gid].pw_len;
575 if (combs_mode == COMBINATOR_MODE_BASE_RIGHT)
577 switch_buffer_by_offset (wordl0, wordl1, wordl2, wordl3, combs_buf[0].pw_len);
580 const u32 search[4] =
582 digests_buf[digests_offset].digest_buf[DGST_R0],
583 digests_buf[digests_offset].digest_buf[DGST_R1],
584 digests_buf[digests_offset].digest_buf[DGST_R2],
585 digests_buf[digests_offset].digest_buf[DGST_R3]
594 salt_buf[0] = salt_bufs[salt_pos].salt_buf[0];
595 salt_buf[1] = salt_bufs[salt_pos].salt_buf[1];
596 salt_buf[2] = salt_bufs[salt_pos].salt_buf[2];
597 salt_buf[3] = salt_bufs[salt_pos].salt_buf[3];
598 salt_buf[4] = salt_bufs[salt_pos].salt_buf[4];
599 salt_buf[5] = salt_bufs[salt_pos].salt_buf[5];
600 salt_buf[6] = salt_bufs[salt_pos].salt_buf[6];
601 salt_buf[7] = salt_bufs[salt_pos].salt_buf[7];
603 const u32 salt_len = salt_bufs[salt_pos].salt_len;
609 for (u32 il_pos = 0; il_pos < combs_cnt; il_pos++)
611 const u32 pw_r_len = combs_buf[il_pos].pw_len;
613 const u32 pw_len = pw_l_len + pw_r_len;
617 wordr0[0] = combs_buf[il_pos].i[0];
618 wordr0[1] = combs_buf[il_pos].i[1];
619 wordr0[2] = combs_buf[il_pos].i[2];
620 wordr0[3] = combs_buf[il_pos].i[3];
624 wordr1[0] = combs_buf[il_pos].i[4];
625 wordr1[1] = combs_buf[il_pos].i[5];
626 wordr1[2] = combs_buf[il_pos].i[6];
627 wordr1[3] = combs_buf[il_pos].i[7];
643 if (combs_mode == COMBINATOR_MODE_BASE_LEFT)
645 switch_buffer_by_offset (wordr0, wordr1, wordr2, wordr3, pw_l_len);
680 switch_buffer_by_offset (s0, s1, s2, s3, pw_len);
682 const u32 pw_salt_len = pw_len + salt_len;
686 w0[0] = wordl0[0] | wordr0[0];
687 w0[1] = wordl0[1] | wordr0[1];
688 w0[2] = wordl0[2] | wordr0[2];
689 w0[3] = wordl0[3] | wordr0[3];
693 w1[0] = wordl1[0] | wordr1[0];
694 w1[1] = wordl1[1] | wordr1[1];
695 w1[2] = wordl1[2] | wordr1[2];
696 w1[3] = wordl1[3] | wordr1[3];
700 w2[0] = wordl2[0] | wordr2[0];
701 w2[1] = wordl2[1] | wordr2[1];
702 w2[2] = wordl2[2] | wordr2[2];
703 w2[3] = wordl2[3] | wordr2[3];
707 w3[0] = wordl3[0] | wordr3[0];
708 w3[1] = wordl3[1] | wordr3[1];
718 final[ 0] = swap_workaround (w0[0] | s0[0]);
719 final[ 1] = swap_workaround (w0[1] | s0[1]);
720 final[ 2] = swap_workaround (w0[2] | s0[2]);
721 final[ 3] = swap_workaround (w0[3] | s0[3]);
722 final[ 4] = swap_workaround (w1[0] | s1[0]);
723 final[ 5] = swap_workaround (w1[1] | s1[1]);
724 final[ 6] = swap_workaround (w1[2] | s1[2]);
725 final[ 7] = swap_workaround (w1[3] | s1[3]);
726 final[ 8] = swap_workaround (w2[0] | s2[0]);
727 final[ 9] = swap_workaround (w2[1] | s2[1]);
728 final[10] = swap_workaround (w2[2] | s2[2]);
729 final[11] = swap_workaround (w2[3] | s2[3]);
730 final[12] = swap_workaround (w3[0] | s3[0]);
731 final[13] = swap_workaround (w3[1] | s3[1]);
733 final[15] = pw_salt_len * 8;
743 sha1_transform (&final[0], &final[4], &final[8], &final[12], digest);
745 // prepare magic array range
747 u32x lengthMagicArray = 0x20;
748 u32x offsetMagicArray = 0;
750 lengthMagicArray += ((digest[0] >> 24) & 0xff) % 6;
751 lengthMagicArray += ((digest[0] >> 16) & 0xff) % 6;
752 lengthMagicArray += ((digest[0] >> 8) & 0xff) % 6;
753 lengthMagicArray += ((digest[0] >> 0) & 0xff) % 6;
754 lengthMagicArray += ((digest[1] >> 24) & 0xff) % 6;
755 lengthMagicArray += ((digest[1] >> 16) & 0xff) % 6;
756 lengthMagicArray += ((digest[1] >> 8) & 0xff) % 6;
757 lengthMagicArray += ((digest[1] >> 0) & 0xff) % 6;
758 lengthMagicArray += ((digest[2] >> 24) & 0xff) % 6;
759 lengthMagicArray += ((digest[2] >> 16) & 0xff) % 6;
760 offsetMagicArray += ((digest[2] >> 8) & 0xff) % 8;
761 offsetMagicArray += ((digest[2] >> 0) & 0xff) % 8;
762 offsetMagicArray += ((digest[3] >> 24) & 0xff) % 8;
763 offsetMagicArray += ((digest[3] >> 16) & 0xff) % 8;
764 offsetMagicArray += ((digest[3] >> 8) & 0xff) % 8;
765 offsetMagicArray += ((digest[3] >> 0) & 0xff) % 8;
766 offsetMagicArray += ((digest[4] >> 24) & 0xff) % 8;
767 offsetMagicArray += ((digest[4] >> 16) & 0xff) % 8;
768 offsetMagicArray += ((digest[4] >> 8) & 0xff) % 8;
769 offsetMagicArray += ((digest[4] >> 0) & 0xff) % 8;
780 for (int i = 0; i < 64; i++) final[i] = 0;
791 u32 final_len = pw_len;
797 for (i = 0; i < lengthMagicArray - 4; i += 4)
799 const u32 tmp = GETSHIFTEDINT (theMagicArray, offsetMagicArray + i);
801 SETSHIFTEDINT (final, final_len + i, tmp);
804 const u32 mask = 0xffffffff >> (((i - lengthMagicArray) & 3) * 8);
806 const u32 tmp = GETSHIFTEDINT (theMagicArray, offsetMagicArray + i) & mask;
808 SETSHIFTEDINT (final, final_len + i, tmp);
810 final_len += lengthMagicArray;
814 for (i = 0; i < salt_len + 1; i += 4) // +1 for the 0x80
816 const u32 tmp = salt_buf[i / 4]; // attention, int[] not char[]
818 SETSHIFTEDINT (final, final_len + i, tmp);
821 final_len += salt_len;
828 for (left = final_len, off = 0; left >= 56; left -= 64, off += 16)
830 swap_buffer (&final[off]);
832 sha1_transform (&final[off + 0], &final[off + 4], &final[off + 8], &final[off + 12], digest);
835 swap_buffer (&final[off]);
838 final[off + 15] = final_len * 8;
840 sha1_transform (&final[off + 0], &final[off + 4], &final[off + 8], &final[off + 12], digest);
842 const u32x r0 = digest[3];
843 const u32x r1 = digest[4];
844 const u32x r2 = digest[2];
845 const u32x r3 = digest[1];
847 #include VECT_COMPARE_S
851 __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m07800_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)
855 __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m07800_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)