2 * Author......: Jens Steube <jens.steube@gmail.com>
8 //incompatible data-dependant code
9 //#define NEW_SIMD_CODE
11 #include "inc_vendor.cl"
12 #include "inc_hash_constants.h"
13 #include "inc_hash_functions.cl"
14 #include "inc_types.cl"
15 #include "inc_common.cl"
16 #include "inc_simd.cl"
18 #define GETSHIFTEDINT(a,n) amd_bytealign ((a)[((n)/4)+1], (a)[((n)/4)+0], (n))
20 #define SETSHIFTEDINT(a,n,v) \
22 const u32 s = ((n) & 3) * 8; \
23 const u64 x = (u64) (v) << s; \
24 (a)[((n)/4)+0] |= x; \
25 (a)[((n)/4)+1] = x >> 32; \
28 __constant u32 theMagicArray[64] =
30 0x1451ac91,0x4354679f,0xe03be724,0xc27b7428,0xeb133386,0x5ccb4f5a,0x37730a08,0x2f1c5d0e,
31 0xe5e68f33,0xddae9bf8,0x8d4bf216,0xdcd4e12c,0x9ddfcbb0,0x176d70d4,0x3f424df9,0x94111b9b,
32 0x9bc15b9f,0x039d0506,0x8a135e9d,0xe86a9a1e,0x17147cd9,0xf62ac758,0x0a6399a1,0xc370fdd7,
33 0x13745ef6,0x040bc903,0x26f79826,0x2593928a,0x230da2b0,0x6d7963ed,0x3cfa3213,0xa39a0235,
34 0x0a8eddb3,0xc351bf24,0x9f55cd7c,0x4c94af37,0x82520829,0x374e3bb2,0x9107179f,0xcdfd3b11,
35 0, 0, 0, 0, 0, 0, 0, 0,
36 0, 0, 0, 0, 0, 0, 0, 0,
37 0, 0, 0, 0, 0, 0, 0, 0
40 void swap_buffer (u32 final[16])
42 final[ 0] = swap32 (final[ 0]);
43 final[ 1] = swap32 (final[ 1]);
44 final[ 2] = swap32 (final[ 2]);
45 final[ 3] = swap32 (final[ 3]);
46 final[ 4] = swap32 (final[ 4]);
47 final[ 5] = swap32 (final[ 5]);
48 final[ 6] = swap32 (final[ 6]);
49 final[ 7] = swap32 (final[ 7]);
50 final[ 8] = swap32 (final[ 8]);
51 final[ 9] = swap32 (final[ 9]);
52 final[10] = swap32 (final[10]);
53 final[11] = swap32 (final[11]);
54 final[12] = swap32 (final[12]);
55 final[13] = swap32 (final[13]);
56 final[14] = swap32 (final[14]);
57 final[15] = swap32 (final[15]);
60 void sha1_transform (const u32 w0[4], const u32 w1[4], const u32 w2[4], const u32 w3[4], u32 digest[5])
88 SHA1_STEP (SHA1_F0o, A, B, C, D, E, w0_t);
89 SHA1_STEP (SHA1_F0o, E, A, B, C, D, w1_t);
90 SHA1_STEP (SHA1_F0o, D, E, A, B, C, w2_t);
91 SHA1_STEP (SHA1_F0o, C, D, E, A, B, w3_t);
92 SHA1_STEP (SHA1_F0o, B, C, D, E, A, w4_t);
93 SHA1_STEP (SHA1_F0o, A, B, C, D, E, w5_t);
94 SHA1_STEP (SHA1_F0o, E, A, B, C, D, w6_t);
95 SHA1_STEP (SHA1_F0o, D, E, A, B, C, w7_t);
96 SHA1_STEP (SHA1_F0o, C, D, E, A, B, w8_t);
97 SHA1_STEP (SHA1_F0o, B, C, D, E, A, w9_t);
98 SHA1_STEP (SHA1_F0o, A, B, C, D, E, wa_t);
99 SHA1_STEP (SHA1_F0o, E, A, B, C, D, wb_t);
100 SHA1_STEP (SHA1_F0o, D, E, A, B, C, wc_t);
101 SHA1_STEP (SHA1_F0o, C, D, E, A, B, wd_t);
102 SHA1_STEP (SHA1_F0o, B, C, D, E, A, we_t);
103 SHA1_STEP (SHA1_F0o, A, B, C, D, E, wf_t);
104 w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F0o, E, A, B, C, D, w0_t);
105 w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F0o, D, E, A, B, C, w1_t);
106 w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F0o, C, D, E, A, B, w2_t);
107 w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F0o, B, C, D, E, A, w3_t);
112 w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w4_t);
113 w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w5_t);
114 w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w6_t);
115 w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w7_t);
116 w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w8_t);
117 w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w9_t);
118 wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wa_t);
119 wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, wb_t);
120 wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, wc_t);
121 wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, wd_t);
122 we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, we_t);
123 wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wf_t);
124 w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w0_t);
125 w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w1_t);
126 w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w2_t);
127 w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w3_t);
128 w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w4_t);
129 w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w5_t);
130 w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w6_t);
131 w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w7_t);
136 w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, w8_t);
137 w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, w9_t);
138 wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, wa_t);
139 wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, wb_t);
140 wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, wc_t);
141 wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, wd_t);
142 we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, we_t);
143 wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, wf_t);
144 w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, w0_t);
145 w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, w1_t);
146 w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, w2_t);
147 w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, w3_t);
148 w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, w4_t);
149 w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, w5_t);
150 w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, w6_t);
151 w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, w7_t);
152 w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, w8_t);
153 w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, w9_t);
154 wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, wa_t);
155 wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, wb_t);
160 wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, wc_t);
161 wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wd_t);
162 we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, we_t);
163 wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, wf_t);
164 w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w0_t);
165 w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w1_t);
166 w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w2_t);
167 w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w3_t);
168 w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w4_t);
169 w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w5_t);
170 w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w6_t);
171 w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w7_t);
172 w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w8_t);
173 w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w9_t);
174 wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, wa_t);
175 wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, wb_t);
176 wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wc_t);
177 wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, wd_t);
178 we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, we_t);
179 wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, wf_t);
188 void m07800m (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], const u32 pw_len, __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_scryptV0_buf, __global u32 *d_scryptV1_buf, __global u32 *d_scryptV2_buf, __global u32 *d_scryptV3_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)
194 const u32 gid = get_global_id (0);
195 const u32 lid = get_local_id (0);
203 salt_buf[0] = salt_bufs[salt_pos].salt_buf[0];
204 salt_buf[1] = salt_bufs[salt_pos].salt_buf[1];
205 salt_buf[2] = salt_bufs[salt_pos].salt_buf[2];
206 salt_buf[3] = salt_bufs[salt_pos].salt_buf[3];
207 salt_buf[4] = salt_bufs[salt_pos].salt_buf[4];
208 salt_buf[5] = salt_bufs[salt_pos].salt_buf[5];
209 salt_buf[6] = salt_bufs[salt_pos].salt_buf[6];
210 salt_buf[7] = salt_bufs[salt_pos].salt_buf[7];
212 const u32 salt_len = salt_bufs[salt_pos].salt_len;
219 s0[0] = swap32 (salt_buf[0]);
220 s0[1] = swap32 (salt_buf[1]);
221 s0[2] = swap32 (salt_buf[2]);
222 s0[3] = swap32 (salt_buf[3]);
223 s1[0] = swap32 (salt_buf[4]);
224 s1[1] = swap32 (salt_buf[5]);
225 s1[2] = swap32 (salt_buf[6]);
226 s1[3] = swap32 (salt_buf[7]);
236 switch_buffer_by_offset_be_S (s0, s1, s2, s3, pw_len);
238 const u32x pw_salt_len = pw_len + salt_len;
246 for (u32 il_pos = 0; il_pos < il_cnt; il_pos += VECT_SIZE)
248 const u32x w0r = ix_create_bft (bfs_buf, il_pos);
250 const u32x w0lr = w0l | w0r;
260 final[ 0] = w0[0] | s0[0];
261 final[ 1] = w0[1] | s0[1];
262 final[ 2] = w0[2] | s0[2];
263 final[ 3] = w0[3] | s0[3];
264 final[ 4] = w1[0] | s1[0];
265 final[ 5] = w1[1] | s1[1];
266 final[ 6] = w1[2] | s1[2];
267 final[ 7] = w1[3] | s1[3];
268 final[ 8] = w2[0] | s2[0];
269 final[ 9] = w2[1] | s2[1];
270 final[10] = w2[2] | s2[2];
271 final[11] = w2[3] | s2[3];
272 final[12] = w3[0] | s3[0];
273 final[13] = w3[1] | s3[1];
275 final[15] = pw_salt_len * 8;
285 sha1_transform (&final[0], &final[4], &final[8], &final[12], digest);
287 // prepare magic array range
289 u32 lengthMagicArray = 0x20;
290 u32 offsetMagicArray = 0;
292 lengthMagicArray += ((digest[0] >> 24) & 0xff) % 6;
293 lengthMagicArray += ((digest[0] >> 16) & 0xff) % 6;
294 lengthMagicArray += ((digest[0] >> 8) & 0xff) % 6;
295 lengthMagicArray += ((digest[0] >> 0) & 0xff) % 6;
296 lengthMagicArray += ((digest[1] >> 24) & 0xff) % 6;
297 lengthMagicArray += ((digest[1] >> 16) & 0xff) % 6;
298 lengthMagicArray += ((digest[1] >> 8) & 0xff) % 6;
299 lengthMagicArray += ((digest[1] >> 0) & 0xff) % 6;
300 lengthMagicArray += ((digest[2] >> 24) & 0xff) % 6;
301 lengthMagicArray += ((digest[2] >> 16) & 0xff) % 6;
302 offsetMagicArray += ((digest[2] >> 8) & 0xff) % 8;
303 offsetMagicArray += ((digest[2] >> 0) & 0xff) % 8;
304 offsetMagicArray += ((digest[3] >> 24) & 0xff) % 8;
305 offsetMagicArray += ((digest[3] >> 16) & 0xff) % 8;
306 offsetMagicArray += ((digest[3] >> 8) & 0xff) % 8;
307 offsetMagicArray += ((digest[3] >> 0) & 0xff) % 8;
308 offsetMagicArray += ((digest[4] >> 24) & 0xff) % 8;
309 offsetMagicArray += ((digest[4] >> 16) & 0xff) % 8;
310 offsetMagicArray += ((digest[4] >> 8) & 0xff) % 8;
311 offsetMagicArray += ((digest[4] >> 0) & 0xff) % 8;
324 for (int i = 0; i < 32; i++) final[i] = 0;
326 final[0] = swap32 (w0[0]);
327 final[1] = swap32 (w0[1]);
328 final[2] = swap32 (w0[2]);
329 final[3] = swap32 (w0[3]);
330 final[4] = swap32 (w1[0]);
331 final[5] = swap32 (w1[1]);
332 final[6] = swap32 (w1[2]);
333 final[7] = swap32 (w1[3]);
335 u32 final_len = pw_len;
341 for (i = 0; i < lengthMagicArray - 4; i += 4)
343 const u32 tmp = GETSHIFTEDINT (theMagicArray, offsetMagicArray + i);
345 SETSHIFTEDINT (final, final_len + i, tmp);
348 const u32 mask = 0xffffffff >> (((i - lengthMagicArray) & 3) * 8);
350 const u32 tmp = GETSHIFTEDINT (theMagicArray, offsetMagicArray + i) & mask;
352 SETSHIFTEDINT (final, final_len + i, tmp);
354 final_len += lengthMagicArray;
358 for (i = 0; i < salt_len + 1; i += 4) // +1 for the 0x80
360 const u32 tmp = salt_buf[i / 4]; // attention, int[] not char[]
362 SETSHIFTEDINT (final, final_len + i, tmp);
365 final_len += salt_len;
372 for (left = final_len, off = 0; left >= 56; left -= 64, off += 16)
374 swap_buffer (&final[off]);
376 sha1_transform (&final[off + 0], &final[off + 4], &final[off + 8], &final[off + 12], digest);
379 swap_buffer (&final[off]);
382 final[off + 15] = final_len * 8;
384 sha1_transform (&final[off + 0], &final[off + 4], &final[off + 8], &final[off + 12], digest);
386 COMPARE_M_SIMD (digest[3], digest[4], digest[2], digest[1]);
390 void m07800s (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], const u32 pw_len, __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_scryptV0_buf, __global u32 *d_scryptV1_buf, __global u32 *d_scryptV2_buf, __global u32 *d_scryptV3_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)
396 const u32 gid = get_global_id (0);
397 const u32 lid = get_local_id (0);
405 salt_buf[0] = salt_bufs[salt_pos].salt_buf[0];
406 salt_buf[1] = salt_bufs[salt_pos].salt_buf[1];
407 salt_buf[2] = salt_bufs[salt_pos].salt_buf[2];
408 salt_buf[3] = salt_bufs[salt_pos].salt_buf[3];
409 salt_buf[4] = salt_bufs[salt_pos].salt_buf[4];
410 salt_buf[5] = salt_bufs[salt_pos].salt_buf[5];
411 salt_buf[6] = salt_bufs[salt_pos].salt_buf[6];
412 salt_buf[7] = salt_bufs[salt_pos].salt_buf[7];
414 const u32 salt_len = salt_bufs[salt_pos].salt_len;
421 s0[0] = swap32 (salt_buf[0]);
422 s0[1] = swap32 (salt_buf[1]);
423 s0[2] = swap32 (salt_buf[2]);
424 s0[3] = swap32 (salt_buf[3]);
425 s1[0] = swap32 (salt_buf[4]);
426 s1[1] = swap32 (salt_buf[5]);
427 s1[2] = swap32 (salt_buf[6]);
428 s1[3] = swap32 (salt_buf[7]);
438 switch_buffer_by_offset_be_S (s0, s1, s2, s3, pw_len);
440 const u32x pw_salt_len = pw_len + salt_len;
446 const u32 search[4] =
448 digests_buf[digests_offset].digest_buf[DGST_R0],
449 digests_buf[digests_offset].digest_buf[DGST_R1],
450 digests_buf[digests_offset].digest_buf[DGST_R2],
451 digests_buf[digests_offset].digest_buf[DGST_R3]
460 for (u32 il_pos = 0; il_pos < il_cnt; il_pos += VECT_SIZE)
462 const u32x w0r = ix_create_bft (bfs_buf, il_pos);
464 const u32x w0lr = w0l | w0r;
474 final[ 0] = w0[0] | s0[0];
475 final[ 1] = w0[1] | s0[1];
476 final[ 2] = w0[2] | s0[2];
477 final[ 3] = w0[3] | s0[3];
478 final[ 4] = w1[0] | s1[0];
479 final[ 5] = w1[1] | s1[1];
480 final[ 6] = w1[2] | s1[2];
481 final[ 7] = w1[3] | s1[3];
482 final[ 8] = w2[0] | s2[0];
483 final[ 9] = w2[1] | s2[1];
484 final[10] = w2[2] | s2[2];
485 final[11] = w2[3] | s2[3];
486 final[12] = w3[0] | s3[0];
487 final[13] = w3[1] | s3[1];
489 final[15] = pw_salt_len * 8;
499 sha1_transform (&final[0], &final[4], &final[8], &final[12], digest);
501 // prepare magic array range
503 u32 lengthMagicArray = 0x20;
504 u32 offsetMagicArray = 0;
506 lengthMagicArray += ((digest[0] >> 24) & 0xff) % 6;
507 lengthMagicArray += ((digest[0] >> 16) & 0xff) % 6;
508 lengthMagicArray += ((digest[0] >> 8) & 0xff) % 6;
509 lengthMagicArray += ((digest[0] >> 0) & 0xff) % 6;
510 lengthMagicArray += ((digest[1] >> 24) & 0xff) % 6;
511 lengthMagicArray += ((digest[1] >> 16) & 0xff) % 6;
512 lengthMagicArray += ((digest[1] >> 8) & 0xff) % 6;
513 lengthMagicArray += ((digest[1] >> 0) & 0xff) % 6;
514 lengthMagicArray += ((digest[2] >> 24) & 0xff) % 6;
515 lengthMagicArray += ((digest[2] >> 16) & 0xff) % 6;
516 offsetMagicArray += ((digest[2] >> 8) & 0xff) % 8;
517 offsetMagicArray += ((digest[2] >> 0) & 0xff) % 8;
518 offsetMagicArray += ((digest[3] >> 24) & 0xff) % 8;
519 offsetMagicArray += ((digest[3] >> 16) & 0xff) % 8;
520 offsetMagicArray += ((digest[3] >> 8) & 0xff) % 8;
521 offsetMagicArray += ((digest[3] >> 0) & 0xff) % 8;
522 offsetMagicArray += ((digest[4] >> 24) & 0xff) % 8;
523 offsetMagicArray += ((digest[4] >> 16) & 0xff) % 8;
524 offsetMagicArray += ((digest[4] >> 8) & 0xff) % 8;
525 offsetMagicArray += ((digest[4] >> 0) & 0xff) % 8;
538 for (int i = 0; i < 32; i++) final[i] = 0;
540 final[0] = swap32 (w0[0]);
541 final[1] = swap32 (w0[1]);
542 final[2] = swap32 (w0[2]);
543 final[3] = swap32 (w0[3]);
544 final[4] = swap32 (w1[0]);
545 final[5] = swap32 (w1[1]);
546 final[6] = swap32 (w1[2]);
547 final[7] = swap32 (w1[3]);
549 u32 final_len = pw_len;
555 for (i = 0; i < lengthMagicArray - 4; i += 4)
557 const u32 tmp = GETSHIFTEDINT (theMagicArray, offsetMagicArray + i);
559 SETSHIFTEDINT (final, final_len + i, tmp);
562 const u32 mask = 0xffffffff >> (((i - lengthMagicArray) & 3) * 8);
564 const u32 tmp = GETSHIFTEDINT (theMagicArray, offsetMagicArray + i) & mask;
566 SETSHIFTEDINT (final, final_len + i, tmp);
568 final_len += lengthMagicArray;
572 for (i = 0; i < salt_len + 1; i += 4) // +1 for the 0x80
574 const u32 tmp = salt_buf[i / 4]; // attention, int[] not char[]
576 SETSHIFTEDINT (final, final_len + i, tmp);
579 final_len += salt_len;
586 for (left = final_len, off = 0; left >= 56; left -= 64, off += 16)
588 swap_buffer (&final[off]);
590 sha1_transform (&final[off + 0], &final[off + 4], &final[off + 8], &final[off + 12], digest);
593 swap_buffer (&final[off]);
596 final[off + 15] = final_len * 8;
598 sha1_transform (&final[off + 0], &final[off + 4], &final[off + 8], &final[off + 12], digest);
600 COMPARE_S_SIMD (digest[3], digest[4], digest[2], digest[1]);
604 __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_scryptV0_buf, __global u32 *d_scryptV1_buf, __global u32 *d_scryptV2_buf, __global u32 *d_scryptV3_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)
610 const u32 gid = get_global_id (0);
612 if (gid >= gid_max) return;
618 //const u32 lid = get_local_id (0);
622 w0[0] = pws[gid].i[ 0];
623 w0[1] = pws[gid].i[ 1];
624 w0[2] = pws[gid].i[ 2];
625 w0[3] = pws[gid].i[ 3];
648 const u32 pw_len = pws[gid].pw_len;
654 m07800m (w0, w1, w2, w3, pw_len, pws, rules_buf, combs_buf, bfs_buf, tmps, hooks, bitmaps_buf_s1_a, bitmaps_buf_s1_b, bitmaps_buf_s1_c, bitmaps_buf_s1_d, bitmaps_buf_s2_a, bitmaps_buf_s2_b, bitmaps_buf_s2_c, bitmaps_buf_s2_d, plains_buf, digests_buf, hashes_shown, salt_bufs, esalt_bufs, d_return_buf, d_scryptV0_buf, d_scryptV1_buf, d_scryptV2_buf, d_scryptV3_buf, bitmap_mask, bitmap_shift1, bitmap_shift2, salt_pos, loop_pos, loop_cnt, il_cnt, digests_cnt, digests_offset);
657 __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_scryptV0_buf, __global u32 *d_scryptV1_buf, __global u32 *d_scryptV2_buf, __global u32 *d_scryptV3_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)
663 const u32 gid = get_global_id (0);
665 if (gid >= gid_max) return;
671 //const u32 lid = get_local_id (0);
675 w0[0] = pws[gid].i[ 0];
676 w0[1] = pws[gid].i[ 1];
677 w0[2] = pws[gid].i[ 2];
678 w0[3] = pws[gid].i[ 3];
682 w1[0] = pws[gid].i[ 4];
683 w1[1] = pws[gid].i[ 5];
684 w1[2] = pws[gid].i[ 6];
685 w1[3] = pws[gid].i[ 7];
701 const u32 pw_len = pws[gid].pw_len;
707 m07800m (w0, w1, w2, w3, pw_len, pws, rules_buf, combs_buf, bfs_buf, tmps, hooks, bitmaps_buf_s1_a, bitmaps_buf_s1_b, bitmaps_buf_s1_c, bitmaps_buf_s1_d, bitmaps_buf_s2_a, bitmaps_buf_s2_b, bitmaps_buf_s2_c, bitmaps_buf_s2_d, plains_buf, digests_buf, hashes_shown, salt_bufs, esalt_bufs, d_return_buf, d_scryptV0_buf, d_scryptV1_buf, d_scryptV2_buf, d_scryptV3_buf, bitmap_mask, bitmap_shift1, bitmap_shift2, salt_pos, loop_pos, loop_cnt, il_cnt, digests_cnt, digests_offset);
710 __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_scryptV0_buf, __global u32 *d_scryptV1_buf, __global u32 *d_scryptV2_buf, __global u32 *d_scryptV3_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)
714 __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_scryptV0_buf, __global u32 *d_scryptV1_buf, __global u32 *d_scryptV2_buf, __global u32 *d_scryptV3_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)
720 const u32 gid = get_global_id (0);
722 if (gid >= gid_max) return;
728 //const u32 lid = get_local_id (0);
732 w0[0] = pws[gid].i[ 0];
733 w0[1] = pws[gid].i[ 1];
734 w0[2] = pws[gid].i[ 2];
735 w0[3] = pws[gid].i[ 3];
758 const u32 pw_len = pws[gid].pw_len;
764 m07800s (w0, w1, w2, w3, pw_len, pws, rules_buf, combs_buf, bfs_buf, tmps, hooks, bitmaps_buf_s1_a, bitmaps_buf_s1_b, bitmaps_buf_s1_c, bitmaps_buf_s1_d, bitmaps_buf_s2_a, bitmaps_buf_s2_b, bitmaps_buf_s2_c, bitmaps_buf_s2_d, plains_buf, digests_buf, hashes_shown, salt_bufs, esalt_bufs, d_return_buf, d_scryptV0_buf, d_scryptV1_buf, d_scryptV2_buf, d_scryptV3_buf, bitmap_mask, bitmap_shift1, bitmap_shift2, salt_pos, loop_pos, loop_cnt, il_cnt, digests_cnt, digests_offset);
767 __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_scryptV0_buf, __global u32 *d_scryptV1_buf, __global u32 *d_scryptV2_buf, __global u32 *d_scryptV3_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)
773 const u32 gid = get_global_id (0);
775 if (gid >= gid_max) return;
781 //const u32 lid = get_local_id (0);
785 w0[0] = pws[gid].i[ 0];
786 w0[1] = pws[gid].i[ 1];
787 w0[2] = pws[gid].i[ 2];
788 w0[3] = pws[gid].i[ 3];
792 w1[0] = pws[gid].i[ 4];
793 w1[1] = pws[gid].i[ 5];
794 w1[2] = pws[gid].i[ 6];
795 w1[3] = pws[gid].i[ 7];
811 const u32 pw_len = pws[gid].pw_len;
817 m07800s (w0, w1, w2, w3, pw_len, pws, rules_buf, combs_buf, bfs_buf, tmps, hooks, bitmaps_buf_s1_a, bitmaps_buf_s1_b, bitmaps_buf_s1_c, bitmaps_buf_s1_d, bitmaps_buf_s2_a, bitmaps_buf_s2_b, bitmaps_buf_s2_c, bitmaps_buf_s2_d, plains_buf, digests_buf, hashes_shown, salt_bufs, esalt_bufs, d_return_buf, d_scryptV0_buf, d_scryptV1_buf, d_scryptV2_buf, d_scryptV3_buf, bitmap_mask, bitmap_shift1, bitmap_shift2, salt_pos, loop_pos, loop_cnt, il_cnt, digests_cnt, digests_offset);
820 __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_scryptV0_buf, __global u32 *d_scryptV1_buf, __global u32 *d_scryptV2_buf, __global u32 *d_scryptV3_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)