2 * Author......: Jens Steube <jens.steube@gmail.com>
8 //incompatible data-dependant code
9 //#define NEW_SIMD_CODE
11 #include "include/constants.h"
12 #include "include/kernel_vendor.h"
19 #include "include/kernel_functions.c"
20 #include "OpenCL/types_ocl.c"
21 #include "OpenCL/common.c"
22 #include "OpenCL/simd.c"
24 #define GETSHIFTEDINT(a,n) amd_bytealign ((a)[((n)/4)+1], (a)[((n)/4)+0], (n))
26 #define SETSHIFTEDINT(a,n,v) \
28 const u32 s = ((n) & 3) * 8; \
29 const u64 x = (u64) (v) << s; \
30 (a)[((n)/4)+0] |= x; \
31 (a)[((n)/4)+1] = x >> 32; \
34 __constant u32 theMagicArray[64] =
36 0x1451ac91,0x4354679f,0xe03be724,0xc27b7428,0xeb133386,0x5ccb4f5a,0x37730a08,0x2f1c5d0e,
37 0xe5e68f33,0xddae9bf8,0x8d4bf216,0xdcd4e12c,0x9ddfcbb0,0x176d70d4,0x3f424df9,0x94111b9b,
38 0x9bc15b9f,0x039d0506,0x8a135e9d,0xe86a9a1e,0x17147cd9,0xf62ac758,0x0a6399a1,0xc370fdd7,
39 0x13745ef6,0x040bc903,0x26f79826,0x2593928a,0x230da2b0,0x6d7963ed,0x3cfa3213,0xa39a0235,
40 0x0a8eddb3,0xc351bf24,0x9f55cd7c,0x4c94af37,0x82520829,0x374e3bb2,0x9107179f,0xcdfd3b11,
41 0, 0, 0, 0, 0, 0, 0, 0,
42 0, 0, 0, 0, 0, 0, 0, 0,
43 0, 0, 0, 0, 0, 0, 0, 0
46 void swap_buffer (u32 final[16])
48 final[ 0] = swap32 (final[ 0]);
49 final[ 1] = swap32 (final[ 1]);
50 final[ 2] = swap32 (final[ 2]);
51 final[ 3] = swap32 (final[ 3]);
52 final[ 4] = swap32 (final[ 4]);
53 final[ 5] = swap32 (final[ 5]);
54 final[ 6] = swap32 (final[ 6]);
55 final[ 7] = swap32 (final[ 7]);
56 final[ 8] = swap32 (final[ 8]);
57 final[ 9] = swap32 (final[ 9]);
58 final[10] = swap32 (final[10]);
59 final[11] = swap32 (final[11]);
60 final[12] = swap32 (final[12]);
61 final[13] = swap32 (final[13]);
62 final[14] = swap32 (final[14]);
63 final[15] = swap32 (final[15]);
66 void sha1_transform (const u32 w0[4], const u32 w1[4], const u32 w2[4], const u32 w3[4], u32 digest[5])
94 SHA1_STEP (SHA1_F0o, A, B, C, D, E, w0_t);
95 SHA1_STEP (SHA1_F0o, E, A, B, C, D, w1_t);
96 SHA1_STEP (SHA1_F0o, D, E, A, B, C, w2_t);
97 SHA1_STEP (SHA1_F0o, C, D, E, A, B, w3_t);
98 SHA1_STEP (SHA1_F0o, B, C, D, E, A, w4_t);
99 SHA1_STEP (SHA1_F0o, A, B, C, D, E, w5_t);
100 SHA1_STEP (SHA1_F0o, E, A, B, C, D, w6_t);
101 SHA1_STEP (SHA1_F0o, D, E, A, B, C, w7_t);
102 SHA1_STEP (SHA1_F0o, C, D, E, A, B, w8_t);
103 SHA1_STEP (SHA1_F0o, B, C, D, E, A, w9_t);
104 SHA1_STEP (SHA1_F0o, A, B, C, D, E, wa_t);
105 SHA1_STEP (SHA1_F0o, E, A, B, C, D, wb_t);
106 SHA1_STEP (SHA1_F0o, D, E, A, B, C, wc_t);
107 SHA1_STEP (SHA1_F0o, C, D, E, A, B, wd_t);
108 SHA1_STEP (SHA1_F0o, B, C, D, E, A, we_t);
109 SHA1_STEP (SHA1_F0o, A, B, C, D, E, wf_t);
110 w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F0o, E, A, B, C, D, w0_t);
111 w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F0o, D, E, A, B, C, w1_t);
112 w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F0o, C, D, E, A, B, w2_t);
113 w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F0o, B, C, D, E, A, w3_t);
118 w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w4_t);
119 w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w5_t);
120 w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w6_t);
121 w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w7_t);
122 w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w8_t);
123 w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w9_t);
124 wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wa_t);
125 wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, wb_t);
126 wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, wc_t);
127 wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, wd_t);
128 we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, we_t);
129 wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wf_t);
130 w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w0_t);
131 w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w1_t);
132 w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w2_t);
133 w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w3_t);
134 w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w4_t);
135 w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w5_t);
136 w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w6_t);
137 w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w7_t);
142 w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, w8_t);
143 w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, w9_t);
144 wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, wa_t);
145 wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, wb_t);
146 wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, wc_t);
147 wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, wd_t);
148 we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, we_t);
149 wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, wf_t);
150 w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, w0_t);
151 w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, w1_t);
152 w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, w2_t);
153 w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, w3_t);
154 w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, w4_t);
155 w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, w5_t);
156 w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, w6_t);
157 w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, w7_t);
158 w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, w8_t);
159 w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, w9_t);
160 wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, wa_t);
161 wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, wb_t);
166 wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, wc_t);
167 wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wd_t);
168 we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, we_t);
169 wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, wf_t);
170 w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w0_t);
171 w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w1_t);
172 w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w2_t);
173 w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w3_t);
174 w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w4_t);
175 w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w5_t);
176 w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w6_t);
177 w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w7_t);
178 w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w8_t);
179 w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w9_t);
180 wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, wa_t);
181 wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, wb_t);
182 wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wc_t);
183 wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, wd_t);
184 we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, we_t);
185 wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, wf_t);
194 __kernel void m07800_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)
200 const u32 lid = get_local_id (0);
206 const u32 gid = get_global_id (0);
208 if (gid >= gid_max) return;
213 pw_buf0[0] = pws[gid].i[0];
214 pw_buf0[1] = pws[gid].i[1];
215 pw_buf0[2] = pws[gid].i[2];
216 pw_buf0[3] = pws[gid].i[3];
217 pw_buf1[0] = pws[gid].i[4];
218 pw_buf1[1] = pws[gid].i[5];
219 pw_buf1[2] = pws[gid].i[6];
220 pw_buf1[3] = pws[gid].i[7];
222 const u32 pw_l_len = pws[gid].pw_len;
230 salt_buf[0] = salt_bufs[salt_pos].salt_buf[0];
231 salt_buf[1] = salt_bufs[salt_pos].salt_buf[1];
232 salt_buf[2] = salt_bufs[salt_pos].salt_buf[2];
233 salt_buf[3] = salt_bufs[salt_pos].salt_buf[3];
234 salt_buf[4] = salt_bufs[salt_pos].salt_buf[4];
235 salt_buf[5] = salt_bufs[salt_pos].salt_buf[5];
236 salt_buf[6] = salt_bufs[salt_pos].salt_buf[6];
237 salt_buf[7] = salt_bufs[salt_pos].salt_buf[7];
239 const u32 salt_len = salt_bufs[salt_pos].salt_len;
245 for (u32 il_pos = 0; il_pos < il_cnt; il_pos += VECT_SIZE)
247 const u32x pw_r_len = pwlenx_create_combt (combs_buf, il_pos);
249 const u32x pw_len = pw_l_len + pw_r_len;
252 * concat password candidate
255 u32x wordl0[4] = { 0 };
256 u32x wordl1[4] = { 0 };
257 u32x wordl2[4] = { 0 };
258 u32x wordl3[4] = { 0 };
260 wordl0[0] = pw_buf0[0];
261 wordl0[1] = pw_buf0[1];
262 wordl0[2] = pw_buf0[2];
263 wordl0[3] = pw_buf0[3];
264 wordl1[0] = pw_buf1[0];
265 wordl1[1] = pw_buf1[1];
266 wordl1[2] = pw_buf1[2];
267 wordl1[3] = pw_buf1[3];
269 u32x wordr0[4] = { 0 };
270 u32x wordr1[4] = { 0 };
271 u32x wordr2[4] = { 0 };
272 u32x wordr3[4] = { 0 };
274 wordr0[0] = ix_create_combt (combs_buf, il_pos, 0);
275 wordr0[1] = ix_create_combt (combs_buf, il_pos, 1);
276 wordr0[2] = ix_create_combt (combs_buf, il_pos, 2);
277 wordr0[3] = ix_create_combt (combs_buf, il_pos, 3);
278 wordr1[0] = ix_create_combt (combs_buf, il_pos, 4);
279 wordr1[1] = ix_create_combt (combs_buf, il_pos, 5);
280 wordr1[2] = ix_create_combt (combs_buf, il_pos, 6);
281 wordr1[3] = ix_create_combt (combs_buf, il_pos, 7);
283 if (combs_mode == COMBINATOR_MODE_BASE_LEFT)
285 switch_buffer_by_offset_le_VV (wordr0, wordr1, wordr2, wordr3, pw_l_len);
289 switch_buffer_by_offset_le_VV (wordl0, wordl1, wordl2, wordl3, pw_r_len);
297 w0[0] = wordl0[0] | wordr0[0];
298 w0[1] = wordl0[1] | wordr0[1];
299 w0[2] = wordl0[2] | wordr0[2];
300 w0[3] = wordl0[3] | wordr0[3];
301 w1[0] = wordl1[0] | wordr1[0];
302 w1[1] = wordl1[1] | wordr1[1];
303 w1[2] = wordl1[2] | wordr1[2];
304 w1[3] = wordl1[3] | wordr1[3];
305 w2[0] = wordl2[0] | wordr2[0];
306 w2[1] = wordl2[1] | wordr2[1];
307 w2[2] = wordl2[2] | wordr2[2];
308 w2[3] = wordl2[3] | wordr2[3];
309 w3[0] = wordl3[0] | wordr3[0];
310 w3[1] = wordl3[1] | wordr3[1];
311 w3[2] = wordl3[2] | wordr3[2];
312 w3[3] = wordl3[3] | wordr3[3];
340 switch_buffer_by_offset_le_VV (s0, s1, s2, s3, pw_len);
342 const u32x pw_salt_len = pw_len + salt_len;
350 final[ 0] = swap32 (w0[0] | s0[0]);
351 final[ 1] = swap32 (w0[1] | s0[1]);
352 final[ 2] = swap32 (w0[2] | s0[2]);
353 final[ 3] = swap32 (w0[3] | s0[3]);
354 final[ 4] = swap32 (w1[0] | s1[0]);
355 final[ 5] = swap32 (w1[1] | s1[1]);
356 final[ 6] = swap32 (w1[2] | s1[2]);
357 final[ 7] = swap32 (w1[3] | s1[3]);
358 final[ 8] = swap32 (w2[0] | s2[0]);
359 final[ 9] = swap32 (w2[1] | s2[1]);
360 final[10] = swap32 (w2[2] | s2[2]);
361 final[11] = swap32 (w2[3] | s2[3]);
362 final[12] = swap32 (w3[0] | s3[0]);
363 final[13] = swap32 (w3[1] | s3[1]);
365 final[15] = pw_salt_len * 8;
375 sha1_transform (&final[0], &final[4], &final[8], &final[12], digest);
377 // prepare magic array range
379 u32 lengthMagicArray = 0x20;
380 u32 offsetMagicArray = 0;
382 lengthMagicArray += ((digest[0] >> 24) & 0xff) % 6;
383 lengthMagicArray += ((digest[0] >> 16) & 0xff) % 6;
384 lengthMagicArray += ((digest[0] >> 8) & 0xff) % 6;
385 lengthMagicArray += ((digest[0] >> 0) & 0xff) % 6;
386 lengthMagicArray += ((digest[1] >> 24) & 0xff) % 6;
387 lengthMagicArray += ((digest[1] >> 16) & 0xff) % 6;
388 lengthMagicArray += ((digest[1] >> 8) & 0xff) % 6;
389 lengthMagicArray += ((digest[1] >> 0) & 0xff) % 6;
390 lengthMagicArray += ((digest[2] >> 24) & 0xff) % 6;
391 lengthMagicArray += ((digest[2] >> 16) & 0xff) % 6;
392 offsetMagicArray += ((digest[2] >> 8) & 0xff) % 8;
393 offsetMagicArray += ((digest[2] >> 0) & 0xff) % 8;
394 offsetMagicArray += ((digest[3] >> 24) & 0xff) % 8;
395 offsetMagicArray += ((digest[3] >> 16) & 0xff) % 8;
396 offsetMagicArray += ((digest[3] >> 8) & 0xff) % 8;
397 offsetMagicArray += ((digest[3] >> 0) & 0xff) % 8;
398 offsetMagicArray += ((digest[4] >> 24) & 0xff) % 8;
399 offsetMagicArray += ((digest[4] >> 16) & 0xff) % 8;
400 offsetMagicArray += ((digest[4] >> 8) & 0xff) % 8;
401 offsetMagicArray += ((digest[4] >> 0) & 0xff) % 8;
414 for (int i = 0; i < 32; i++) final[i] = 0;
425 u32 final_len = pw_len;
431 for (i = 0; i < lengthMagicArray - 4; i += 4)
433 const u32 tmp = GETSHIFTEDINT (theMagicArray, offsetMagicArray + i);
435 SETSHIFTEDINT (final, final_len + i, tmp);
438 const u32 mask = 0xffffffff >> (((i - lengthMagicArray) & 3) * 8);
440 const u32 tmp = GETSHIFTEDINT (theMagicArray, offsetMagicArray + i) & mask;
442 SETSHIFTEDINT (final, final_len + i, tmp);
444 final_len += lengthMagicArray;
448 for (i = 0; i < salt_len + 1; i += 4) // +1 for the 0x80
450 const u32 tmp = salt_buf[i / 4]; // attention, int[] not char[]
452 SETSHIFTEDINT (final, final_len + i, tmp);
455 final_len += salt_len;
462 for (left = final_len, off = 0; left >= 56; left -= 64, off += 16)
464 swap_buffer (&final[off]);
466 sha1_transform (&final[off + 0], &final[off + 4], &final[off + 8], &final[off + 12], digest);
469 swap_buffer (&final[off]);
472 final[off + 15] = final_len * 8;
474 sha1_transform (&final[off + 0], &final[off + 4], &final[off + 8], &final[off + 12], digest);
476 COMPARE_M_SIMD (digest[3], digest[4], digest[2], digest[1]);
480 __kernel void m07800_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)
484 __kernel void m07800_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)
488 __kernel void m07800_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)
494 const u32 lid = get_local_id (0);
500 const u32 gid = get_global_id (0);
502 if (gid >= gid_max) return;
507 pw_buf0[0] = pws[gid].i[0];
508 pw_buf0[1] = pws[gid].i[1];
509 pw_buf0[2] = pws[gid].i[2];
510 pw_buf0[3] = pws[gid].i[3];
511 pw_buf1[0] = pws[gid].i[4];
512 pw_buf1[1] = pws[gid].i[5];
513 pw_buf1[2] = pws[gid].i[6];
514 pw_buf1[3] = pws[gid].i[7];
516 const u32 pw_l_len = pws[gid].pw_len;
524 salt_buf[0] = salt_bufs[salt_pos].salt_buf[0];
525 salt_buf[1] = salt_bufs[salt_pos].salt_buf[1];
526 salt_buf[2] = salt_bufs[salt_pos].salt_buf[2];
527 salt_buf[3] = salt_bufs[salt_pos].salt_buf[3];
528 salt_buf[4] = salt_bufs[salt_pos].salt_buf[4];
529 salt_buf[5] = salt_bufs[salt_pos].salt_buf[5];
530 salt_buf[6] = salt_bufs[salt_pos].salt_buf[6];
531 salt_buf[7] = salt_bufs[salt_pos].salt_buf[7];
533 const u32 salt_len = salt_bufs[salt_pos].salt_len;
539 const u32 search[4] =
541 digests_buf[digests_offset].digest_buf[DGST_R0],
542 digests_buf[digests_offset].digest_buf[DGST_R1],
543 digests_buf[digests_offset].digest_buf[DGST_R2],
544 digests_buf[digests_offset].digest_buf[DGST_R3]
551 for (u32 il_pos = 0; il_pos < il_cnt; il_pos += VECT_SIZE)
553 const u32x pw_r_len = pwlenx_create_combt (combs_buf, il_pos);
555 const u32x pw_len = pw_l_len + pw_r_len;
558 * concat password candidate
561 u32x wordl0[4] = { 0 };
562 u32x wordl1[4] = { 0 };
563 u32x wordl2[4] = { 0 };
564 u32x wordl3[4] = { 0 };
566 wordl0[0] = pw_buf0[0];
567 wordl0[1] = pw_buf0[1];
568 wordl0[2] = pw_buf0[2];
569 wordl0[3] = pw_buf0[3];
570 wordl1[0] = pw_buf1[0];
571 wordl1[1] = pw_buf1[1];
572 wordl1[2] = pw_buf1[2];
573 wordl1[3] = pw_buf1[3];
575 u32x wordr0[4] = { 0 };
576 u32x wordr1[4] = { 0 };
577 u32x wordr2[4] = { 0 };
578 u32x wordr3[4] = { 0 };
580 wordr0[0] = ix_create_combt (combs_buf, il_pos, 0);
581 wordr0[1] = ix_create_combt (combs_buf, il_pos, 1);
582 wordr0[2] = ix_create_combt (combs_buf, il_pos, 2);
583 wordr0[3] = ix_create_combt (combs_buf, il_pos, 3);
584 wordr1[0] = ix_create_combt (combs_buf, il_pos, 4);
585 wordr1[1] = ix_create_combt (combs_buf, il_pos, 5);
586 wordr1[2] = ix_create_combt (combs_buf, il_pos, 6);
587 wordr1[3] = ix_create_combt (combs_buf, il_pos, 7);
589 if (combs_mode == COMBINATOR_MODE_BASE_LEFT)
591 switch_buffer_by_offset_le_VV (wordr0, wordr1, wordr2, wordr3, pw_l_len);
595 switch_buffer_by_offset_le_VV (wordl0, wordl1, wordl2, wordl3, pw_r_len);
603 w0[0] = wordl0[0] | wordr0[0];
604 w0[1] = wordl0[1] | wordr0[1];
605 w0[2] = wordl0[2] | wordr0[2];
606 w0[3] = wordl0[3] | wordr0[3];
607 w1[0] = wordl1[0] | wordr1[0];
608 w1[1] = wordl1[1] | wordr1[1];
609 w1[2] = wordl1[2] | wordr1[2];
610 w1[3] = wordl1[3] | wordr1[3];
611 w2[0] = wordl2[0] | wordr2[0];
612 w2[1] = wordl2[1] | wordr2[1];
613 w2[2] = wordl2[2] | wordr2[2];
614 w2[3] = wordl2[3] | wordr2[3];
615 w3[0] = wordl3[0] | wordr3[0];
616 w3[1] = wordl3[1] | wordr3[1];
617 w3[2] = wordl3[2] | wordr3[2];
618 w3[3] = wordl3[3] | wordr3[3];
646 switch_buffer_by_offset_le_VV (s0, s1, s2, s3, pw_len);
648 const u32x pw_salt_len = pw_len + salt_len;
656 final[ 0] = swap32 (w0[0] | s0[0]);
657 final[ 1] = swap32 (w0[1] | s0[1]);
658 final[ 2] = swap32 (w0[2] | s0[2]);
659 final[ 3] = swap32 (w0[3] | s0[3]);
660 final[ 4] = swap32 (w1[0] | s1[0]);
661 final[ 5] = swap32 (w1[1] | s1[1]);
662 final[ 6] = swap32 (w1[2] | s1[2]);
663 final[ 7] = swap32 (w1[3] | s1[3]);
664 final[ 8] = swap32 (w2[0] | s2[0]);
665 final[ 9] = swap32 (w2[1] | s2[1]);
666 final[10] = swap32 (w2[2] | s2[2]);
667 final[11] = swap32 (w2[3] | s2[3]);
668 final[12] = swap32 (w3[0] | s3[0]);
669 final[13] = swap32 (w3[1] | s3[1]);
671 final[15] = pw_salt_len * 8;
681 sha1_transform (&final[0], &final[4], &final[8], &final[12], digest);
683 // prepare magic array range
685 u32 lengthMagicArray = 0x20;
686 u32 offsetMagicArray = 0;
688 lengthMagicArray += ((digest[0] >> 24) & 0xff) % 6;
689 lengthMagicArray += ((digest[0] >> 16) & 0xff) % 6;
690 lengthMagicArray += ((digest[0] >> 8) & 0xff) % 6;
691 lengthMagicArray += ((digest[0] >> 0) & 0xff) % 6;
692 lengthMagicArray += ((digest[1] >> 24) & 0xff) % 6;
693 lengthMagicArray += ((digest[1] >> 16) & 0xff) % 6;
694 lengthMagicArray += ((digest[1] >> 8) & 0xff) % 6;
695 lengthMagicArray += ((digest[1] >> 0) & 0xff) % 6;
696 lengthMagicArray += ((digest[2] >> 24) & 0xff) % 6;
697 lengthMagicArray += ((digest[2] >> 16) & 0xff) % 6;
698 offsetMagicArray += ((digest[2] >> 8) & 0xff) % 8;
699 offsetMagicArray += ((digest[2] >> 0) & 0xff) % 8;
700 offsetMagicArray += ((digest[3] >> 24) & 0xff) % 8;
701 offsetMagicArray += ((digest[3] >> 16) & 0xff) % 8;
702 offsetMagicArray += ((digest[3] >> 8) & 0xff) % 8;
703 offsetMagicArray += ((digest[3] >> 0) & 0xff) % 8;
704 offsetMagicArray += ((digest[4] >> 24) & 0xff) % 8;
705 offsetMagicArray += ((digest[4] >> 16) & 0xff) % 8;
706 offsetMagicArray += ((digest[4] >> 8) & 0xff) % 8;
707 offsetMagicArray += ((digest[4] >> 0) & 0xff) % 8;
720 for (int i = 0; i < 32; i++) final[i] = 0;
731 u32 final_len = pw_len;
737 for (i = 0; i < lengthMagicArray - 4; i += 4)
739 const u32 tmp = GETSHIFTEDINT (theMagicArray, offsetMagicArray + i);
741 SETSHIFTEDINT (final, final_len + i, tmp);
744 const u32 mask = 0xffffffff >> (((i - lengthMagicArray) & 3) * 8);
746 const u32 tmp = GETSHIFTEDINT (theMagicArray, offsetMagicArray + i) & mask;
748 SETSHIFTEDINT (final, final_len + i, tmp);
750 final_len += lengthMagicArray;
754 for (i = 0; i < salt_len + 1; i += 4) // +1 for the 0x80
756 const u32 tmp = salt_buf[i / 4]; // attention, int[] not char[]
758 SETSHIFTEDINT (final, final_len + i, tmp);
761 final_len += salt_len;
768 for (left = final_len, off = 0; left >= 56; left -= 64, off += 16)
770 swap_buffer (&final[off]);
772 sha1_transform (&final[off + 0], &final[off + 4], &final[off + 8], &final[off + 12], digest);
775 swap_buffer (&final[off]);
778 final[off + 15] = final_len * 8;
780 sha1_transform (&final[off + 0], &final[off + 4], &final[off + 8], &final[off + 12], digest);
782 COMPARE_S_SIMD (digest[3], digest[4], digest[2], digest[1]);
786 __kernel void m07800_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)
790 __kernel void m07800_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)