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"
19 #include "include/rp_gpu.h"
22 #define COMPARE_S "check_single_comp4.c"
23 #define COMPARE_M "check_multi_comp4.c"
25 __constant u64 k_sha512[80] =
27 SHA512C00, SHA512C01, SHA512C02, SHA512C03,
28 SHA512C04, SHA512C05, SHA512C06, SHA512C07,
29 SHA512C08, SHA512C09, SHA512C0a, SHA512C0b,
30 SHA512C0c, SHA512C0d, SHA512C0e, SHA512C0f,
31 SHA512C10, SHA512C11, SHA512C12, SHA512C13,
32 SHA512C14, SHA512C15, SHA512C16, SHA512C17,
33 SHA512C18, SHA512C19, SHA512C1a, SHA512C1b,
34 SHA512C1c, SHA512C1d, SHA512C1e, SHA512C1f,
35 SHA512C20, SHA512C21, SHA512C22, SHA512C23,
36 SHA512C24, SHA512C25, SHA512C26, SHA512C27,
37 SHA512C28, SHA512C29, SHA512C2a, SHA512C2b,
38 SHA512C2c, SHA512C2d, SHA512C2e, SHA512C2f,
39 SHA512C30, SHA512C31, SHA512C32, SHA512C33,
40 SHA512C34, SHA512C35, SHA512C36, SHA512C37,
41 SHA512C38, SHA512C39, SHA512C3a, SHA512C3b,
42 SHA512C3c, SHA512C3d, SHA512C3e, SHA512C3f,
43 SHA512C40, SHA512C41, SHA512C42, SHA512C43,
44 SHA512C44, SHA512C45, SHA512C46, SHA512C47,
45 SHA512C48, SHA512C49, SHA512C4a, SHA512C4b,
46 SHA512C4c, SHA512C4d, SHA512C4e, SHA512C4f,
49 static void sha512_transform (const u32 w0[4], const u32 w1[4], const u32 w2[4], const u32 w3[4], u64 digest[8])
51 u64 w0_t = hl32_to_64 (w0[0], w0[1]);
52 u64 w1_t = hl32_to_64 (w0[2], w0[3]);
53 u64 w2_t = hl32_to_64 (w1[0], w1[1]);
54 u64 w3_t = hl32_to_64 (w1[2], w1[3]);
55 u64 w4_t = hl32_to_64 (w2[0], w2[1]);
56 u64 w5_t = hl32_to_64 (w2[2], w2[3]);
57 u64 w6_t = hl32_to_64 (w3[0], w3[1]);
66 u64 wf_t = hl32_to_64 (w3[2], w3[3]);
77 #define ROUND_EXPAND() \
79 w0_t = SHA512_EXPAND (we_t, w9_t, w1_t, w0_t); \
80 w1_t = SHA512_EXPAND (wf_t, wa_t, w2_t, w1_t); \
81 w2_t = SHA512_EXPAND (w0_t, wb_t, w3_t, w2_t); \
82 w3_t = SHA512_EXPAND (w1_t, wc_t, w4_t, w3_t); \
83 w4_t = SHA512_EXPAND (w2_t, wd_t, w5_t, w4_t); \
84 w5_t = SHA512_EXPAND (w3_t, we_t, w6_t, w5_t); \
85 w6_t = SHA512_EXPAND (w4_t, wf_t, w7_t, w6_t); \
86 w7_t = SHA512_EXPAND (w5_t, w0_t, w8_t, w7_t); \
87 w8_t = SHA512_EXPAND (w6_t, w1_t, w9_t, w8_t); \
88 w9_t = SHA512_EXPAND (w7_t, w2_t, wa_t, w9_t); \
89 wa_t = SHA512_EXPAND (w8_t, w3_t, wb_t, wa_t); \
90 wb_t = SHA512_EXPAND (w9_t, w4_t, wc_t, wb_t); \
91 wc_t = SHA512_EXPAND (wa_t, w5_t, wd_t, wc_t); \
92 wd_t = SHA512_EXPAND (wb_t, w6_t, we_t, wd_t); \
93 we_t = SHA512_EXPAND (wc_t, w7_t, wf_t, we_t); \
94 wf_t = SHA512_EXPAND (wd_t, w8_t, w0_t, wf_t); \
97 #define ROUND_STEP(i) \
99 SHA512_STEP (SHA512_F0o, SHA512_F1o, a, b, c, d, e, f, g, h, w0_t, k_sha512[i + 0]); \
100 SHA512_STEP (SHA512_F0o, SHA512_F1o, h, a, b, c, d, e, f, g, w1_t, k_sha512[i + 1]); \
101 SHA512_STEP (SHA512_F0o, SHA512_F1o, g, h, a, b, c, d, e, f, w2_t, k_sha512[i + 2]); \
102 SHA512_STEP (SHA512_F0o, SHA512_F1o, f, g, h, a, b, c, d, e, w3_t, k_sha512[i + 3]); \
103 SHA512_STEP (SHA512_F0o, SHA512_F1o, e, f, g, h, a, b, c, d, w4_t, k_sha512[i + 4]); \
104 SHA512_STEP (SHA512_F0o, SHA512_F1o, d, e, f, g, h, a, b, c, w5_t, k_sha512[i + 5]); \
105 SHA512_STEP (SHA512_F0o, SHA512_F1o, c, d, e, f, g, h, a, b, w6_t, k_sha512[i + 6]); \
106 SHA512_STEP (SHA512_F0o, SHA512_F1o, b, c, d, e, f, g, h, a, w7_t, k_sha512[i + 7]); \
107 SHA512_STEP (SHA512_F0o, SHA512_F1o, a, b, c, d, e, f, g, h, w8_t, k_sha512[i + 8]); \
108 SHA512_STEP (SHA512_F0o, SHA512_F1o, h, a, b, c, d, e, f, g, w9_t, k_sha512[i + 9]); \
109 SHA512_STEP (SHA512_F0o, SHA512_F1o, g, h, a, b, c, d, e, f, wa_t, k_sha512[i + 10]); \
110 SHA512_STEP (SHA512_F0o, SHA512_F1o, f, g, h, a, b, c, d, e, wb_t, k_sha512[i + 11]); \
111 SHA512_STEP (SHA512_F0o, SHA512_F1o, e, f, g, h, a, b, c, d, wc_t, k_sha512[i + 12]); \
112 SHA512_STEP (SHA512_F0o, SHA512_F1o, d, e, f, g, h, a, b, c, wd_t, k_sha512[i + 13]); \
113 SHA512_STEP (SHA512_F0o, SHA512_F1o, c, d, e, f, g, h, a, b, we_t, k_sha512[i + 14]); \
114 SHA512_STEP (SHA512_F0o, SHA512_F1o, b, c, d, e, f, g, h, a, wf_t, k_sha512[i + 15]); \
120 for (int i = 16; i < 80; i += 16)
122 ROUND_EXPAND (); ROUND_STEP (i);
146 __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m01740_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 rules_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
152 const u32 lid = get_local_id (0);
158 const u32 gid = get_global_id (0);
160 if (gid >= gid_max) return;
164 pw_buf0[0] = pws[gid].i[ 0];
165 pw_buf0[1] = pws[gid].i[ 1];
166 pw_buf0[2] = pws[gid].i[ 2];
167 pw_buf0[3] = pws[gid].i[ 3];
171 pw_buf1[0] = pws[gid].i[ 4];
172 pw_buf1[1] = pws[gid].i[ 5];
173 pw_buf1[2] = pws[gid].i[ 6];
174 pw_buf1[3] = pws[gid].i[ 7];
176 const u32 pw_len = pws[gid].pw_len;
184 salt_buf0[0] = salt_bufs[salt_pos].salt_buf[0];
185 salt_buf0[1] = salt_bufs[salt_pos].salt_buf[1];
186 salt_buf0[2] = salt_bufs[salt_pos].salt_buf[2];
187 salt_buf0[3] = salt_bufs[salt_pos].salt_buf[3];
191 salt_buf1[0] = salt_bufs[salt_pos].salt_buf[4];
192 salt_buf1[1] = salt_bufs[salt_pos].salt_buf[5];
193 salt_buf1[2] = salt_bufs[salt_pos].salt_buf[6];
194 salt_buf1[3] = salt_bufs[salt_pos].salt_buf[7];
196 const u32 salt_len = salt_bufs[salt_pos].salt_len;
202 for (u32 il_pos = 0; il_pos < rules_cnt; il_pos++)
226 const u32 out_len = apply_rules (rules_buf[il_pos].cmds, w0, w1, pw_len);
232 const u32 out_salt_len = (out_len * 2) + salt_len;
239 make_unicode (w0, w0_t, w1_t);
240 make_unicode (w1, w2_t, w3_t);
242 switch_buffer_by_offset (w0_t, w1_t, w2_t, w3_t, salt_len);
244 w0_t[0] |= salt_buf0[0];
245 w0_t[1] |= salt_buf0[1];
246 w0_t[2] |= salt_buf0[2];
247 w0_t[3] |= salt_buf0[3];
248 w1_t[0] |= salt_buf1[0];
249 w1_t[1] |= salt_buf1[1];
250 w1_t[2] |= salt_buf1[2];
251 w1_t[3] |= salt_buf1[3];
253 append_0x80_4 (w0_t, w1_t, w2_t, w3_t, out_salt_len);
259 w0_t[0] = swap_workaround (w0_t[0]);
260 w0_t[1] = swap_workaround (w0_t[1]);
261 w0_t[2] = swap_workaround (w0_t[2]);
262 w0_t[3] = swap_workaround (w0_t[3]);
263 w1_t[0] = swap_workaround (w1_t[0]);
264 w1_t[1] = swap_workaround (w1_t[1]);
265 w1_t[2] = swap_workaround (w1_t[2]);
266 w1_t[3] = swap_workaround (w1_t[3]);
267 w2_t[0] = swap_workaround (w2_t[0]);
268 w2_t[1] = swap_workaround (w2_t[1]);
269 w2_t[2] = swap_workaround (w2_t[2]);
270 w2_t[3] = swap_workaround (w2_t[3]);
271 w3_t[0] = swap_workaround (w3_t[0]);
272 w3_t[1] = swap_workaround (w3_t[1]);
274 w3_t[3] = out_salt_len * 8;
278 digest[0] = SHA512M_A;
279 digest[1] = SHA512M_B;
280 digest[2] = SHA512M_C;
281 digest[3] = SHA512M_D;
282 digest[4] = SHA512M_E;
283 digest[5] = SHA512M_F;
284 digest[6] = SHA512M_G;
285 digest[7] = SHA512M_H;
287 sha512_transform (w0_t, w1_t, w2_t, w3_t, digest);
290 const u32 r0 = l32_from_64 (digest[7]);
291 const u32 r1 = h32_from_64 (digest[7]);
292 const u32 r2 = l32_from_64 (digest[3]);
293 const u32 r3 = h32_from_64 (digest[3]);
299 __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m01740_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 rules_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
303 __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m01740_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 rules_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
307 __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m01740_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 rules_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
313 const u32 lid = get_local_id (0);
319 const u32 gid = get_global_id (0);
321 if (gid >= gid_max) return;
325 pw_buf0[0] = pws[gid].i[ 0];
326 pw_buf0[1] = pws[gid].i[ 1];
327 pw_buf0[2] = pws[gid].i[ 2];
328 pw_buf0[3] = pws[gid].i[ 3];
332 pw_buf1[0] = pws[gid].i[ 4];
333 pw_buf1[1] = pws[gid].i[ 5];
334 pw_buf1[2] = pws[gid].i[ 6];
335 pw_buf1[3] = pws[gid].i[ 7];
337 const u32 pw_len = pws[gid].pw_len;
345 salt_buf0[0] = salt_bufs[salt_pos].salt_buf[0];
346 salt_buf0[1] = salt_bufs[salt_pos].salt_buf[1];
347 salt_buf0[2] = salt_bufs[salt_pos].salt_buf[2];
348 salt_buf0[3] = salt_bufs[salt_pos].salt_buf[3];
352 salt_buf1[0] = salt_bufs[salt_pos].salt_buf[4];
353 salt_buf1[1] = salt_bufs[salt_pos].salt_buf[5];
354 salt_buf1[2] = salt_bufs[salt_pos].salt_buf[6];
355 salt_buf1[3] = salt_bufs[salt_pos].salt_buf[7];
357 const u32 salt_len = salt_bufs[salt_pos].salt_len;
363 const u32 search[4] =
365 digests_buf[digests_offset].digest_buf[DGST_R0],
366 digests_buf[digests_offset].digest_buf[DGST_R1],
367 digests_buf[digests_offset].digest_buf[DGST_R2],
368 digests_buf[digests_offset].digest_buf[DGST_R3]
375 for (u32 il_pos = 0; il_pos < rules_cnt; il_pos++)
399 const u32 out_len = apply_rules (rules_buf[il_pos].cmds, w0, w1, pw_len);
405 const u32 out_salt_len = (out_len * 2) + salt_len;
412 make_unicode (w0, w0_t, w1_t);
413 make_unicode (w1, w2_t, w3_t);
415 switch_buffer_by_offset (w0_t, w1_t, w2_t, w3_t, salt_len);
417 w0_t[0] |= salt_buf0[0];
418 w0_t[1] |= salt_buf0[1];
419 w0_t[2] |= salt_buf0[2];
420 w0_t[3] |= salt_buf0[3];
421 w1_t[0] |= salt_buf1[0];
422 w1_t[1] |= salt_buf1[1];
423 w1_t[2] |= salt_buf1[2];
424 w1_t[3] |= salt_buf1[3];
426 append_0x80_4 (w0_t, w1_t, w2_t, w3_t, out_salt_len);
432 w0_t[0] = swap_workaround (w0_t[0]);
433 w0_t[1] = swap_workaround (w0_t[1]);
434 w0_t[2] = swap_workaround (w0_t[2]);
435 w0_t[3] = swap_workaround (w0_t[3]);
436 w1_t[0] = swap_workaround (w1_t[0]);
437 w1_t[1] = swap_workaround (w1_t[1]);
438 w1_t[2] = swap_workaround (w1_t[2]);
439 w1_t[3] = swap_workaround (w1_t[3]);
440 w2_t[0] = swap_workaround (w2_t[0]);
441 w2_t[1] = swap_workaround (w2_t[1]);
442 w2_t[2] = swap_workaround (w2_t[2]);
443 w2_t[3] = swap_workaround (w2_t[3]);
444 w3_t[0] = swap_workaround (w3_t[0]);
445 w3_t[1] = swap_workaround (w3_t[1]);
447 w3_t[3] = out_salt_len * 8;
451 digest[0] = SHA512M_A;
452 digest[1] = SHA512M_B;
453 digest[2] = SHA512M_C;
454 digest[3] = SHA512M_D;
455 digest[4] = SHA512M_E;
456 digest[5] = SHA512M_F;
457 digest[6] = SHA512M_G;
458 digest[7] = SHA512M_H;
460 sha512_transform (w0_t, w1_t, w2_t, w3_t, digest);
463 const u32 r0 = l32_from_64 (digest[7]);
464 const u32 r1 = h32_from_64 (digest[7]);
465 const u32 r2 = l32_from_64 (digest[3]);
466 const u32 r3 = h32_from_64 (digest[3]);
472 __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m01740_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 rules_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
476 __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m01740_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 rules_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)