2 * Author......: Jens Steube <jens.steube@gmail.com>
10 #include "inc_hash_constants.h"
11 #include "inc_vendor.cl"
18 #include "inc_hash_functions.cl"
19 #include "inc_types.cl"
20 #include "inc_common.cl"
21 #include "inc_simd.cl"
23 __constant u64 k_sha512[80] =
25 SHA512C00, SHA512C01, SHA512C02, SHA512C03,
26 SHA512C04, SHA512C05, SHA512C06, SHA512C07,
27 SHA512C08, SHA512C09, SHA512C0a, SHA512C0b,
28 SHA512C0c, SHA512C0d, SHA512C0e, SHA512C0f,
29 SHA512C10, SHA512C11, SHA512C12, SHA512C13,
30 SHA512C14, SHA512C15, SHA512C16, SHA512C17,
31 SHA512C18, SHA512C19, SHA512C1a, SHA512C1b,
32 SHA512C1c, SHA512C1d, SHA512C1e, SHA512C1f,
33 SHA512C20, SHA512C21, SHA512C22, SHA512C23,
34 SHA512C24, SHA512C25, SHA512C26, SHA512C27,
35 SHA512C28, SHA512C29, SHA512C2a, SHA512C2b,
36 SHA512C2c, SHA512C2d, SHA512C2e, SHA512C2f,
37 SHA512C30, SHA512C31, SHA512C32, SHA512C33,
38 SHA512C34, SHA512C35, SHA512C36, SHA512C37,
39 SHA512C38, SHA512C39, SHA512C3a, SHA512C3b,
40 SHA512C3c, SHA512C3d, SHA512C3e, SHA512C3f,
41 SHA512C40, SHA512C41, SHA512C42, SHA512C43,
42 SHA512C44, SHA512C45, SHA512C46, SHA512C47,
43 SHA512C48, SHA512C49, SHA512C4a, SHA512C4b,
44 SHA512C4c, SHA512C4d, SHA512C4e, SHA512C4f,
47 void sha512_transform (const u64x w0[4], const u64x w1[4], const u64x w2[4], const u64x w3[4], u64x digest[8])
75 #define ROUND_EXPAND() \
77 w0_t = SHA512_EXPAND (we_t, w9_t, w1_t, w0_t); \
78 w1_t = SHA512_EXPAND (wf_t, wa_t, w2_t, w1_t); \
79 w2_t = SHA512_EXPAND (w0_t, wb_t, w3_t, w2_t); \
80 w3_t = SHA512_EXPAND (w1_t, wc_t, w4_t, w3_t); \
81 w4_t = SHA512_EXPAND (w2_t, wd_t, w5_t, w4_t); \
82 w5_t = SHA512_EXPAND (w3_t, we_t, w6_t, w5_t); \
83 w6_t = SHA512_EXPAND (w4_t, wf_t, w7_t, w6_t); \
84 w7_t = SHA512_EXPAND (w5_t, w0_t, w8_t, w7_t); \
85 w8_t = SHA512_EXPAND (w6_t, w1_t, w9_t, w8_t); \
86 w9_t = SHA512_EXPAND (w7_t, w2_t, wa_t, w9_t); \
87 wa_t = SHA512_EXPAND (w8_t, w3_t, wb_t, wa_t); \
88 wb_t = SHA512_EXPAND (w9_t, w4_t, wc_t, wb_t); \
89 wc_t = SHA512_EXPAND (wa_t, w5_t, wd_t, wc_t); \
90 wd_t = SHA512_EXPAND (wb_t, w6_t, we_t, wd_t); \
91 we_t = SHA512_EXPAND (wc_t, w7_t, wf_t, we_t); \
92 wf_t = SHA512_EXPAND (wd_t, w8_t, w0_t, wf_t); \
95 #define ROUND_STEP(i) \
97 SHA512_STEP (SHA512_F0o, SHA512_F1o, a, b, c, d, e, f, g, h, w0_t, k_sha512[i + 0]); \
98 SHA512_STEP (SHA512_F0o, SHA512_F1o, h, a, b, c, d, e, f, g, w1_t, k_sha512[i + 1]); \
99 SHA512_STEP (SHA512_F0o, SHA512_F1o, g, h, a, b, c, d, e, f, w2_t, k_sha512[i + 2]); \
100 SHA512_STEP (SHA512_F0o, SHA512_F1o, f, g, h, a, b, c, d, e, w3_t, k_sha512[i + 3]); \
101 SHA512_STEP (SHA512_F0o, SHA512_F1o, e, f, g, h, a, b, c, d, w4_t, k_sha512[i + 4]); \
102 SHA512_STEP (SHA512_F0o, SHA512_F1o, d, e, f, g, h, a, b, c, w5_t, k_sha512[i + 5]); \
103 SHA512_STEP (SHA512_F0o, SHA512_F1o, c, d, e, f, g, h, a, b, w6_t, k_sha512[i + 6]); \
104 SHA512_STEP (SHA512_F0o, SHA512_F1o, b, c, d, e, f, g, h, a, w7_t, k_sha512[i + 7]); \
105 SHA512_STEP (SHA512_F0o, SHA512_F1o, a, b, c, d, e, f, g, h, w8_t, k_sha512[i + 8]); \
106 SHA512_STEP (SHA512_F0o, SHA512_F1o, h, a, b, c, d, e, f, g, w9_t, k_sha512[i + 9]); \
107 SHA512_STEP (SHA512_F0o, SHA512_F1o, g, h, a, b, c, d, e, f, wa_t, k_sha512[i + 10]); \
108 SHA512_STEP (SHA512_F0o, SHA512_F1o, f, g, h, a, b, c, d, e, wb_t, k_sha512[i + 11]); \
109 SHA512_STEP (SHA512_F0o, SHA512_F1o, e, f, g, h, a, b, c, d, wc_t, k_sha512[i + 12]); \
110 SHA512_STEP (SHA512_F0o, SHA512_F1o, d, e, f, g, h, a, b, c, wd_t, k_sha512[i + 13]); \
111 SHA512_STEP (SHA512_F0o, SHA512_F1o, c, d, e, f, g, h, a, b, we_t, k_sha512[i + 14]); \
112 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);
135 void hmac_sha512_pad (u32x w0[4], u32x w1[4], u32x w2[4], u32x w3[4], u64x ipad[8], u64x opad[8])
142 w0_t[0] = hl32_to_64 (w0[0], w0[1]) ^ (u64x) 0x3636363636363636;
143 w0_t[1] = hl32_to_64 (w0[2], w0[3]) ^ (u64x) 0x3636363636363636;
144 w0_t[2] = hl32_to_64 (w1[0], w1[1]) ^ (u64x) 0x3636363636363636;
145 w0_t[3] = hl32_to_64 (w1[2], w1[3]) ^ (u64x) 0x3636363636363636;
146 w1_t[0] = hl32_to_64 (w2[0], w2[1]) ^ (u64x) 0x3636363636363636;
147 w1_t[1] = hl32_to_64 (w2[2], w2[3]) ^ (u64x) 0x3636363636363636;
148 w1_t[2] = hl32_to_64 (w3[0], w3[1]) ^ (u64x) 0x3636363636363636;
149 w1_t[3] = hl32_to_64 (w3[2], w3[3]) ^ (u64x) 0x3636363636363636;
150 w2_t[0] = (u64x) 0x3636363636363636;
151 w2_t[1] = (u64x) 0x3636363636363636;
152 w2_t[2] = (u64x) 0x3636363636363636;
153 w2_t[3] = (u64x) 0x3636363636363636;
154 w3_t[0] = (u64x) 0x3636363636363636;
155 w3_t[1] = (u64x) 0x3636363636363636;
156 w3_t[2] = (u64x) 0x3636363636363636;
157 w3_t[3] = (u64x) 0x3636363636363636;
168 sha512_transform (w0_t, w1_t, w2_t, w3_t, ipad);
170 w0_t[0] = hl32_to_64 (w0[0], w0[1]) ^ (u64x) 0x5c5c5c5c5c5c5c5c;
171 w0_t[1] = hl32_to_64 (w0[2], w0[3]) ^ (u64x) 0x5c5c5c5c5c5c5c5c;
172 w0_t[2] = hl32_to_64 (w1[0], w1[1]) ^ (u64x) 0x5c5c5c5c5c5c5c5c;
173 w0_t[3] = hl32_to_64 (w1[2], w1[3]) ^ (u64x) 0x5c5c5c5c5c5c5c5c;
174 w1_t[0] = hl32_to_64 (w2[0], w2[1]) ^ (u64x) 0x5c5c5c5c5c5c5c5c;
175 w1_t[1] = hl32_to_64 (w2[2], w2[3]) ^ (u64x) 0x5c5c5c5c5c5c5c5c;
176 w1_t[2] = hl32_to_64 (w3[0], w3[1]) ^ (u64x) 0x5c5c5c5c5c5c5c5c;
177 w1_t[3] = hl32_to_64 (w3[2], w3[3]) ^ (u64x) 0x5c5c5c5c5c5c5c5c;
178 w2_t[0] = (u64x) 0x5c5c5c5c5c5c5c5c;
179 w2_t[1] = (u64x) 0x5c5c5c5c5c5c5c5c;
180 w2_t[2] = (u64x) 0x5c5c5c5c5c5c5c5c;
181 w2_t[3] = (u64x) 0x5c5c5c5c5c5c5c5c;
182 w3_t[0] = (u64x) 0x5c5c5c5c5c5c5c5c;
183 w3_t[1] = (u64x) 0x5c5c5c5c5c5c5c5c;
184 w3_t[2] = (u64x) 0x5c5c5c5c5c5c5c5c;
185 w3_t[3] = (u64x) 0x5c5c5c5c5c5c5c5c;
196 sha512_transform (w0_t, w1_t, w2_t, w3_t, opad);
199 void hmac_sha512_run (u32x w0[4], u32x w1[4], u32x w2[4], u32x w3[4], u64x ipad[8], u64x opad[8], u64x digest[8])
206 w0_t[0] = hl32_to_64 (w0[0], w0[1]);
207 w0_t[1] = hl32_to_64 (w0[2], w0[3]);
208 w0_t[2] = hl32_to_64 (w1[0], w1[1]);
209 w0_t[3] = hl32_to_64 (w1[2], w1[3]);
210 w1_t[0] = hl32_to_64 (w2[0], w2[1]);
211 w1_t[1] = hl32_to_64 (w2[2], w2[3]);
212 w1_t[2] = hl32_to_64 (w3[0], w3[1]);
221 w3_t[3] = hl32_to_64 (w3[2], w3[3]);
232 sha512_transform (w0_t, w1_t, w2_t, w3_t, digest);
242 w2_t[0] = 0x8000000000000000;
249 w3_t[3] = (128 + 64) * 8;
260 sha512_transform (w0_t, w1_t, w2_t, w3_t, digest);
263 __kernel void m01760_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_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 il_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
269 const u32 lid = get_local_id (0);
275 const u32 gid = get_global_id (0);
277 if (gid >= gid_max) return;
282 pw_buf0[0] = pws[gid].i[0];
283 pw_buf0[1] = pws[gid].i[1];
284 pw_buf0[2] = pws[gid].i[2];
285 pw_buf0[3] = pws[gid].i[3];
286 pw_buf1[0] = pws[gid].i[4];
287 pw_buf1[1] = pws[gid].i[5];
288 pw_buf1[2] = pws[gid].i[6];
289 pw_buf1[3] = pws[gid].i[7];
291 const u32 pw_l_len = pws[gid].pw_len;
302 salt_buf0[0] = swap32_S (salt_bufs[salt_pos].salt_buf[ 0]);
303 salt_buf0[1] = swap32_S (salt_bufs[salt_pos].salt_buf[ 1]);
304 salt_buf0[2] = swap32_S (salt_bufs[salt_pos].salt_buf[ 2]);
305 salt_buf0[3] = swap32_S (salt_bufs[salt_pos].salt_buf[ 3]);
306 salt_buf1[0] = swap32_S (salt_bufs[salt_pos].salt_buf[ 4]);
307 salt_buf1[1] = swap32_S (salt_bufs[salt_pos].salt_buf[ 5]);
308 salt_buf1[2] = swap32_S (salt_bufs[salt_pos].salt_buf[ 6]);
309 salt_buf1[3] = swap32_S (salt_bufs[salt_pos].salt_buf[ 7]);
310 salt_buf2[0] = swap32_S (salt_bufs[salt_pos].salt_buf[ 8]);
311 salt_buf2[1] = swap32_S (salt_bufs[salt_pos].salt_buf[ 9]);
312 salt_buf2[2] = swap32_S (salt_bufs[salt_pos].salt_buf[10]);
313 salt_buf2[3] = swap32_S (salt_bufs[salt_pos].salt_buf[11]);
314 salt_buf3[0] = swap32_S (salt_bufs[salt_pos].salt_buf[12]);
315 salt_buf3[1] = swap32_S (salt_bufs[salt_pos].salt_buf[13]);
316 salt_buf3[2] = swap32_S (salt_bufs[salt_pos].salt_buf[14]);
317 salt_buf3[3] = swap32_S (salt_bufs[salt_pos].salt_buf[15]);
319 const u32 salt_len = salt_bufs[salt_pos].salt_len;
330 w0_t[0] = salt_buf0[0];
331 w0_t[1] = salt_buf0[1];
332 w0_t[2] = salt_buf0[2];
333 w0_t[3] = salt_buf0[3];
334 w1_t[0] = salt_buf1[0];
335 w1_t[1] = salt_buf1[1];
336 w1_t[2] = salt_buf1[2];
337 w1_t[3] = salt_buf1[3];
338 w2_t[0] = salt_buf2[0];
339 w2_t[1] = salt_buf2[1];
340 w2_t[2] = salt_buf2[2];
341 w2_t[3] = salt_buf2[3];
342 w3_t[0] = salt_buf3[0];
343 w3_t[1] = salt_buf3[1];
344 w3_t[2] = salt_buf3[2];
345 w3_t[3] = salt_buf3[3];
350 hmac_sha512_pad (w0_t, w1_t, w2_t, w3_t, ipad, opad);
356 for (u32 il_pos = 0; il_pos < il_cnt; il_pos += VECT_SIZE)
358 const u32x pw_r_len = pwlenx_create_combt (combs_buf, il_pos);
360 const u32x pw_len = pw_l_len + pw_r_len;
363 * concat password candidate
366 u32x wordl0[4] = { 0 };
367 u32x wordl1[4] = { 0 };
368 u32x wordl2[4] = { 0 };
369 u32x wordl3[4] = { 0 };
371 wordl0[0] = pw_buf0[0];
372 wordl0[1] = pw_buf0[1];
373 wordl0[2] = pw_buf0[2];
374 wordl0[3] = pw_buf0[3];
375 wordl1[0] = pw_buf1[0];
376 wordl1[1] = pw_buf1[1];
377 wordl1[2] = pw_buf1[2];
378 wordl1[3] = pw_buf1[3];
380 u32x wordr0[4] = { 0 };
381 u32x wordr1[4] = { 0 };
382 u32x wordr2[4] = { 0 };
383 u32x wordr3[4] = { 0 };
385 wordr0[0] = ix_create_combt (combs_buf, il_pos, 0);
386 wordr0[1] = ix_create_combt (combs_buf, il_pos, 1);
387 wordr0[2] = ix_create_combt (combs_buf, il_pos, 2);
388 wordr0[3] = ix_create_combt (combs_buf, il_pos, 3);
389 wordr1[0] = ix_create_combt (combs_buf, il_pos, 4);
390 wordr1[1] = ix_create_combt (combs_buf, il_pos, 5);
391 wordr1[2] = ix_create_combt (combs_buf, il_pos, 6);
392 wordr1[3] = ix_create_combt (combs_buf, il_pos, 7);
394 if (combs_mode == COMBINATOR_MODE_BASE_LEFT)
396 switch_buffer_by_offset_le_VV (wordr0, wordr1, wordr2, wordr3, pw_l_len);
400 switch_buffer_by_offset_le_VV (wordl0, wordl1, wordl2, wordl3, pw_r_len);
408 w0[0] = wordl0[0] | wordr0[0];
409 w0[1] = wordl0[1] | wordr0[1];
410 w0[2] = wordl0[2] | wordr0[2];
411 w0[3] = wordl0[3] | wordr0[3];
412 w1[0] = wordl1[0] | wordr1[0];
413 w1[1] = wordl1[1] | wordr1[1];
414 w1[2] = wordl1[2] | wordr1[2];
415 w1[3] = wordl1[3] | wordr1[3];
416 w2[0] = wordl2[0] | wordr2[0];
417 w2[1] = wordl2[1] | wordr2[1];
418 w2[2] = wordl2[2] | wordr2[2];
419 w2[3] = wordl2[3] | wordr2[3];
420 w3[0] = wordl3[0] | wordr3[0];
421 w3[1] = wordl3[1] | wordr3[1];
422 w3[2] = wordl3[2] | wordr3[2];
423 w3[3] = wordl3[3] | wordr3[3];
425 append_0x80_4x4_VV (w0, w1, w2, w3, pw_len);
427 w0[0] = swap32 (w0[0]);
428 w0[1] = swap32 (w0[1]);
429 w0[2] = swap32 (w0[2]);
430 w0[3] = swap32 (w0[3]);
431 w1[0] = swap32 (w1[0]);
432 w1[1] = swap32 (w1[1]);
433 w1[2] = swap32 (w1[2]);
434 w1[3] = swap32 (w1[3]);
435 w2[0] = swap32 (w2[0]);
436 w2[1] = swap32 (w2[1]);
437 w2[2] = swap32 (w2[2]);
438 w2[3] = swap32 (w2[3]);
439 w3[0] = swap32 (w3[0]);
440 w3[1] = swap32 (w3[1]);
442 w3[3] = (128 + pw_len) * 8;
446 hmac_sha512_run (w0, w1, w2, w3, ipad, opad, digest);
448 const u32x r0 = l32_from_64 (digest[7]);
449 const u32x r1 = h32_from_64 (digest[7]);
450 const u32x r2 = l32_from_64 (digest[3]);
451 const u32x r3 = h32_from_64 (digest[3]);
453 COMPARE_M_SIMD (r0, r1, r2, r3);
457 __kernel void m01760_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_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 il_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
461 __kernel void m01760_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_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 il_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
465 __kernel void m01760_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_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 il_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
471 const u32 lid = get_local_id (0);
477 const u32 gid = get_global_id (0);
479 if (gid >= gid_max) return;
484 pw_buf0[0] = pws[gid].i[0];
485 pw_buf0[1] = pws[gid].i[1];
486 pw_buf0[2] = pws[gid].i[2];
487 pw_buf0[3] = pws[gid].i[3];
488 pw_buf1[0] = pws[gid].i[4];
489 pw_buf1[1] = pws[gid].i[5];
490 pw_buf1[2] = pws[gid].i[6];
491 pw_buf1[3] = pws[gid].i[7];
493 const u32 pw_l_len = pws[gid].pw_len;
504 salt_buf0[0] = swap32_S (salt_bufs[salt_pos].salt_buf[ 0]);
505 salt_buf0[1] = swap32_S (salt_bufs[salt_pos].salt_buf[ 1]);
506 salt_buf0[2] = swap32_S (salt_bufs[salt_pos].salt_buf[ 2]);
507 salt_buf0[3] = swap32_S (salt_bufs[salt_pos].salt_buf[ 3]);
508 salt_buf1[0] = swap32_S (salt_bufs[salt_pos].salt_buf[ 4]);
509 salt_buf1[1] = swap32_S (salt_bufs[salt_pos].salt_buf[ 5]);
510 salt_buf1[2] = swap32_S (salt_bufs[salt_pos].salt_buf[ 6]);
511 salt_buf1[3] = swap32_S (salt_bufs[salt_pos].salt_buf[ 7]);
512 salt_buf2[0] = swap32_S (salt_bufs[salt_pos].salt_buf[ 8]);
513 salt_buf2[1] = swap32_S (salt_bufs[salt_pos].salt_buf[ 9]);
514 salt_buf2[2] = swap32_S (salt_bufs[salt_pos].salt_buf[10]);
515 salt_buf2[3] = swap32_S (salt_bufs[salt_pos].salt_buf[11]);
516 salt_buf3[0] = swap32_S (salt_bufs[salt_pos].salt_buf[12]);
517 salt_buf3[1] = swap32_S (salt_bufs[salt_pos].salt_buf[13]);
518 salt_buf3[2] = swap32_S (salt_bufs[salt_pos].salt_buf[14]);
519 salt_buf3[3] = swap32_S (salt_bufs[salt_pos].salt_buf[15]);
521 const u32 salt_len = salt_bufs[salt_pos].salt_len;
532 w0_t[0] = salt_buf0[0];
533 w0_t[1] = salt_buf0[1];
534 w0_t[2] = salt_buf0[2];
535 w0_t[3] = salt_buf0[3];
536 w1_t[0] = salt_buf1[0];
537 w1_t[1] = salt_buf1[1];
538 w1_t[2] = salt_buf1[2];
539 w1_t[3] = salt_buf1[3];
540 w2_t[0] = salt_buf2[0];
541 w2_t[1] = salt_buf2[1];
542 w2_t[2] = salt_buf2[2];
543 w2_t[3] = salt_buf2[3];
544 w3_t[0] = salt_buf3[0];
545 w3_t[1] = salt_buf3[1];
546 w3_t[2] = salt_buf3[2];
547 w3_t[3] = salt_buf3[3];
552 hmac_sha512_pad (w0_t, w1_t, w2_t, w3_t, ipad, opad);
558 const u32 search[4] =
560 digests_buf[digests_offset].digest_buf[DGST_R0],
561 digests_buf[digests_offset].digest_buf[DGST_R1],
562 digests_buf[digests_offset].digest_buf[DGST_R2],
563 digests_buf[digests_offset].digest_buf[DGST_R3]
570 for (u32 il_pos = 0; il_pos < il_cnt; il_pos += VECT_SIZE)
572 const u32x pw_r_len = pwlenx_create_combt (combs_buf, il_pos);
574 const u32x pw_len = pw_l_len + pw_r_len;
577 * concat password candidate
580 u32x wordl0[4] = { 0 };
581 u32x wordl1[4] = { 0 };
582 u32x wordl2[4] = { 0 };
583 u32x wordl3[4] = { 0 };
585 wordl0[0] = pw_buf0[0];
586 wordl0[1] = pw_buf0[1];
587 wordl0[2] = pw_buf0[2];
588 wordl0[3] = pw_buf0[3];
589 wordl1[0] = pw_buf1[0];
590 wordl1[1] = pw_buf1[1];
591 wordl1[2] = pw_buf1[2];
592 wordl1[3] = pw_buf1[3];
594 u32x wordr0[4] = { 0 };
595 u32x wordr1[4] = { 0 };
596 u32x wordr2[4] = { 0 };
597 u32x wordr3[4] = { 0 };
599 wordr0[0] = ix_create_combt (combs_buf, il_pos, 0);
600 wordr0[1] = ix_create_combt (combs_buf, il_pos, 1);
601 wordr0[2] = ix_create_combt (combs_buf, il_pos, 2);
602 wordr0[3] = ix_create_combt (combs_buf, il_pos, 3);
603 wordr1[0] = ix_create_combt (combs_buf, il_pos, 4);
604 wordr1[1] = ix_create_combt (combs_buf, il_pos, 5);
605 wordr1[2] = ix_create_combt (combs_buf, il_pos, 6);
606 wordr1[3] = ix_create_combt (combs_buf, il_pos, 7);
608 if (combs_mode == COMBINATOR_MODE_BASE_LEFT)
610 switch_buffer_by_offset_le_VV (wordr0, wordr1, wordr2, wordr3, pw_l_len);
614 switch_buffer_by_offset_le_VV (wordl0, wordl1, wordl2, wordl3, pw_r_len);
622 w0[0] = wordl0[0] | wordr0[0];
623 w0[1] = wordl0[1] | wordr0[1];
624 w0[2] = wordl0[2] | wordr0[2];
625 w0[3] = wordl0[3] | wordr0[3];
626 w1[0] = wordl1[0] | wordr1[0];
627 w1[1] = wordl1[1] | wordr1[1];
628 w1[2] = wordl1[2] | wordr1[2];
629 w1[3] = wordl1[3] | wordr1[3];
630 w2[0] = wordl2[0] | wordr2[0];
631 w2[1] = wordl2[1] | wordr2[1];
632 w2[2] = wordl2[2] | wordr2[2];
633 w2[3] = wordl2[3] | wordr2[3];
634 w3[0] = wordl3[0] | wordr3[0];
635 w3[1] = wordl3[1] | wordr3[1];
636 w3[2] = wordl3[2] | wordr3[2];
637 w3[3] = wordl3[3] | wordr3[3];
639 append_0x80_4x4_VV (w0, w1, w2, w3, pw_len);
641 w0[0] = swap32 (w0[0]);
642 w0[1] = swap32 (w0[1]);
643 w0[2] = swap32 (w0[2]);
644 w0[3] = swap32 (w0[3]);
645 w1[0] = swap32 (w1[0]);
646 w1[1] = swap32 (w1[1]);
647 w1[2] = swap32 (w1[2]);
648 w1[3] = swap32 (w1[3]);
649 w2[0] = swap32 (w2[0]);
650 w2[1] = swap32 (w2[1]);
651 w2[2] = swap32 (w2[2]);
652 w2[3] = swap32 (w2[3]);
653 w3[0] = swap32 (w3[0]);
654 w3[1] = swap32 (w3[1]);
656 w3[3] = (128 + pw_len) * 8;
660 hmac_sha512_run (w0, w1, w2, w3, ipad, opad, digest);
662 const u32x r0 = l32_from_64 (digest[7]);
663 const u32x r1 = h32_from_64 (digest[7]);
664 const u32x r2 = l32_from_64 (digest[3]);
665 const u32x r3 = h32_from_64 (digest[3]);
667 COMPARE_S_SIMD (r0, r1, r2, r3);
671 __kernel void m01760_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_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 il_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
675 __kernel void m01760_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_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 il_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)