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 "types_ocl.c"
21 #define COMPARE_M "check_multi_vect1_comp4.c"
25 #define COMPARE_M "check_multi_vect2_comp4.c"
29 #define COMPARE_M "check_multi_vect4_comp4.c"
32 static void sha1_transform (const u32 w0[4], const u32 w1[4], const u32 w2[4], const u32 w3[4], u32 digest[5])
60 SHA1_STEP (SHA1_F0o, A, B, C, D, E, w0_t);
61 SHA1_STEP (SHA1_F0o, E, A, B, C, D, w1_t);
62 SHA1_STEP (SHA1_F0o, D, E, A, B, C, w2_t);
63 SHA1_STEP (SHA1_F0o, C, D, E, A, B, w3_t);
64 SHA1_STEP (SHA1_F0o, B, C, D, E, A, w4_t);
65 SHA1_STEP (SHA1_F0o, A, B, C, D, E, w5_t);
66 SHA1_STEP (SHA1_F0o, E, A, B, C, D, w6_t);
67 SHA1_STEP (SHA1_F0o, D, E, A, B, C, w7_t);
68 SHA1_STEP (SHA1_F0o, C, D, E, A, B, w8_t);
69 SHA1_STEP (SHA1_F0o, B, C, D, E, A, w9_t);
70 SHA1_STEP (SHA1_F0o, A, B, C, D, E, wa_t);
71 SHA1_STEP (SHA1_F0o, E, A, B, C, D, wb_t);
72 SHA1_STEP (SHA1_F0o, D, E, A, B, C, wc_t);
73 SHA1_STEP (SHA1_F0o, C, D, E, A, B, wd_t);
74 SHA1_STEP (SHA1_F0o, B, C, D, E, A, we_t);
75 SHA1_STEP (SHA1_F0o, A, B, C, D, E, wf_t);
76 w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F0o, E, A, B, C, D, w0_t);
77 w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F0o, D, E, A, B, C, w1_t);
78 w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F0o, C, D, E, A, B, w2_t);
79 w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F0o, B, C, D, E, A, w3_t);
84 w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w4_t);
85 w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w5_t);
86 w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w6_t);
87 w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w7_t);
88 w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w8_t);
89 w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w9_t);
90 wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wa_t);
91 wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, wb_t);
92 wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, wc_t);
93 wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, wd_t);
94 we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, we_t);
95 wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wf_t);
96 w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w0_t);
97 w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w1_t);
98 w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w2_t);
99 w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w3_t);
100 w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w4_t);
101 w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w5_t);
102 w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w6_t);
103 w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w7_t);
108 w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, w8_t);
109 w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, w9_t);
110 wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, wa_t);
111 wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, wb_t);
112 wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, wc_t);
113 wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, wd_t);
114 we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, we_t);
115 wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, wf_t);
116 w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, w0_t);
117 w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, w1_t);
118 w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, w2_t);
119 w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, w3_t);
120 w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, w4_t);
121 w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, w5_t);
122 w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, w6_t);
123 w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, w7_t);
124 w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, w8_t);
125 w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, w9_t);
126 wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, wa_t);
127 wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, wb_t);
132 wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, wc_t);
133 wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wd_t);
134 we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, we_t);
135 wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, wf_t);
136 w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w0_t);
137 w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w1_t);
138 w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w2_t);
139 w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w3_t);
140 w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w4_t);
141 w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w5_t);
142 w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w6_t);
143 w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w7_t);
144 w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w8_t);
145 w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w9_t);
146 wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, wa_t);
147 wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, wb_t);
148 wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wc_t);
149 wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, wd_t);
150 we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, we_t);
151 wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, wf_t);
160 __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m10300_init (__global pw_t *pws, __global gpu_rule_t *rules_buf, __global comb_t *combs_buf, __global bf_t *bfs_buf, __global saph_sha1_tmp_t *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 rules_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
166 const u32 gid = get_global_id (0);
168 if (gid >= gid_max) return;
172 word_buf0[0] = pws[gid].i[0];
173 word_buf0[1] = pws[gid].i[1];
174 word_buf0[2] = pws[gid].i[2];
175 word_buf0[3] = pws[gid].i[3];
179 word_buf1[0] = pws[gid].i[4];
180 word_buf1[1] = pws[gid].i[5];
181 word_buf1[2] = pws[gid].i[6];
182 word_buf1[3] = pws[gid].i[7];
186 word_buf2[0] = pws[gid].i[8];
187 word_buf2[1] = pws[gid].i[9];
189 const u32 pw_len = pws[gid].pw_len;
195 u32 salt_len = salt_bufs[salt_pos].salt_len;
199 salt_buf[0] = salt_bufs[salt_pos].salt_buf[0];
200 salt_buf[1] = salt_bufs[salt_pos].salt_buf[1];
201 salt_buf[2] = salt_bufs[salt_pos].salt_buf[2];
202 salt_buf[3] = salt_bufs[salt_pos].salt_buf[3];
236 switch_buffer_by_offset (w0, w1, w2, w3, pw_len);
238 w0[0] |= word_buf0[0];
239 w0[1] |= word_buf0[1];
240 w0[2] |= word_buf0[2];
241 w0[3] |= word_buf0[3];
243 w1[0] |= word_buf1[0];
244 w1[1] |= word_buf1[1];
245 w1[2] |= word_buf1[2];
246 w1[3] |= word_buf1[3];
248 w2[0] |= word_buf2[0];
249 w2[1] |= word_buf2[1];
251 const u32 pw_salt_len = pw_len + salt_len;
253 append_0x80_4 (w0, w1, w2, w3, pw_salt_len);
257 w0[0] = swap_workaround (w0[0]);
258 w0[1] = swap_workaround (w0[1]);
259 w0[2] = swap_workaround (w0[2]);
260 w0[3] = swap_workaround (w0[3]);
262 w1[0] = swap_workaround (w1[0]);
263 w1[1] = swap_workaround (w1[1]);
264 w1[2] = swap_workaround (w1[2]);
265 w1[3] = swap_workaround (w1[3]);
267 w2[0] = swap_workaround (w2[0]);
268 w2[1] = swap_workaround (w2[1]);
269 w2[2] = swap_workaround (w2[2]);
270 w2[3] = swap_workaround (w2[3]);
272 w3[0] = swap_workaround (w3[0]);
273 w3[1] = swap_workaround (w3[1]);
274 w3[2] = swap_workaround (w3[2]);
275 w3[3] = pw_salt_len * 8;
285 sha1_transform (w0, w1, w2, w3, digest);
287 tmps[gid].digest_buf[0] = digest[0];
288 tmps[gid].digest_buf[1] = digest[1];
289 tmps[gid].digest_buf[2] = digest[2];
290 tmps[gid].digest_buf[3] = digest[3];
291 tmps[gid].digest_buf[4] = digest[4];
294 __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m10300_loop (__global pw_t *pws, __global gpu_rule_t *rules_buf, __global comb_t *combs_buf, __global bf_t *bfs_buf, __global saph_sha1_tmp_t *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 rules_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
300 const u32 gid = get_global_id (0);
304 word_buf0[0] = swap_workaround (pws[gid].i[0]);
305 word_buf0[1] = swap_workaround (pws[gid].i[1]);
306 word_buf0[2] = swap_workaround (pws[gid].i[2]);
307 word_buf0[3] = swap_workaround (pws[gid].i[3]);
311 word_buf1[0] = swap_workaround (pws[gid].i[4]);
312 word_buf1[1] = swap_workaround (pws[gid].i[5]);
313 word_buf1[2] = swap_workaround (pws[gid].i[6]);
314 word_buf1[3] = swap_workaround (pws[gid].i[7]);
318 word_buf2[0] = swap_workaround (pws[gid].i[8]);
319 word_buf2[1] = swap_workaround (pws[gid].i[9]);
321 const u32 pw_len = pws[gid].pw_len;
325 digest[0] = tmps[gid].digest_buf[0];
326 digest[1] = tmps[gid].digest_buf[1];
327 digest[2] = tmps[gid].digest_buf[2];
328 digest[3] = tmps[gid].digest_buf[3];
329 digest[4] = tmps[gid].digest_buf[4];
335 for (u32 i = 0; i < loop_cnt; i++)
359 switch_buffer_by_offset_be (w0, w1, w2, w3, pw_len);
361 w0[0] |= word_buf0[0];
362 w0[1] |= word_buf0[1];
363 w0[2] |= word_buf0[2];
364 w0[3] |= word_buf0[3];
365 w1[0] |= word_buf1[0];
366 w1[1] |= word_buf1[1];
367 w1[2] |= word_buf1[2];
368 w1[3] |= word_buf1[3];
369 w2[0] |= word_buf2[0];
370 w2[1] |= word_buf2[1];
375 w3[3] = (pw_len + 20) * 8;
383 sha1_transform (w0, w1, w2, w3, digest);
386 tmps[gid].digest_buf[0] = digest[0];
387 tmps[gid].digest_buf[1] = digest[1];
388 tmps[gid].digest_buf[2] = digest[2];
389 tmps[gid].digest_buf[3] = digest[3];
390 tmps[gid].digest_buf[4] = digest[4];
393 __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m10300_comp (__global pw_t *pws, __global gpu_rule_t *rules_buf, __global comb_t *combs_buf, __global bf_t *bfs_buf, __global saph_sha1_tmp_t *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 rules_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
399 const u32 gid = get_global_id (0);
401 if (gid >= gid_max) return;
403 const u32 lid = get_local_id (0);
409 const u32 r0 = tmps[gid].digest_buf[0];
410 const u32 r1 = tmps[gid].digest_buf[1];
411 const u32 r2 = tmps[gid].digest_buf[2];
412 const u32 r3 = tmps[gid].digest_buf[3];