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 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_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)
200 const u32 gid = get_global_id (0);
201 const u32 lid = get_local_id (0);
209 salt_buf[0] = salt_bufs[salt_pos].salt_buf[0];
210 salt_buf[1] = salt_bufs[salt_pos].salt_buf[1];
211 salt_buf[2] = salt_bufs[salt_pos].salt_buf[2];
212 salt_buf[3] = salt_bufs[salt_pos].salt_buf[3];
213 salt_buf[4] = salt_bufs[salt_pos].salt_buf[4];
214 salt_buf[5] = salt_bufs[salt_pos].salt_buf[5];
215 salt_buf[6] = salt_bufs[salt_pos].salt_buf[6];
216 salt_buf[7] = salt_bufs[salt_pos].salt_buf[7];
218 const u32 salt_len = salt_bufs[salt_pos].salt_len;
225 s0[0] = swap32 (salt_buf[0]);
226 s0[1] = swap32 (salt_buf[1]);
227 s0[2] = swap32 (salt_buf[2]);
228 s0[3] = swap32 (salt_buf[3]);
229 s1[0] = swap32 (salt_buf[4]);
230 s1[1] = swap32 (salt_buf[5]);
231 s1[2] = swap32 (salt_buf[6]);
232 s1[3] = swap32 (salt_buf[7]);
242 switch_buffer_by_offset_be_S (s0, s1, s2, s3, pw_len);
244 const u32x pw_salt_len = pw_len + salt_len;
252 for (u32 il_pos = 0; il_pos < il_cnt; il_pos += VECT_SIZE)
254 const u32x w0r = ix_create_bft (bfs_buf, il_pos);
256 const u32x w0lr = w0l | w0r;
266 final[ 0] = w0[0] | s0[0];
267 final[ 1] = w0[1] | s0[1];
268 final[ 2] = w0[2] | s0[2];
269 final[ 3] = w0[3] | s0[3];
270 final[ 4] = w1[0] | s1[0];
271 final[ 5] = w1[1] | s1[1];
272 final[ 6] = w1[2] | s1[2];
273 final[ 7] = w1[3] | s1[3];
274 final[ 8] = w2[0] | s2[0];
275 final[ 9] = w2[1] | s2[1];
276 final[10] = w2[2] | s2[2];
277 final[11] = w2[3] | s2[3];
278 final[12] = w3[0] | s3[0];
279 final[13] = w3[1] | s3[1];
281 final[15] = pw_salt_len * 8;
291 sha1_transform (&final[0], &final[4], &final[8], &final[12], digest);
293 // prepare magic array range
295 u32 lengthMagicArray = 0x20;
296 u32 offsetMagicArray = 0;
298 lengthMagicArray += ((digest[0] >> 24) & 0xff) % 6;
299 lengthMagicArray += ((digest[0] >> 16) & 0xff) % 6;
300 lengthMagicArray += ((digest[0] >> 8) & 0xff) % 6;
301 lengthMagicArray += ((digest[0] >> 0) & 0xff) % 6;
302 lengthMagicArray += ((digest[1] >> 24) & 0xff) % 6;
303 lengthMagicArray += ((digest[1] >> 16) & 0xff) % 6;
304 lengthMagicArray += ((digest[1] >> 8) & 0xff) % 6;
305 lengthMagicArray += ((digest[1] >> 0) & 0xff) % 6;
306 lengthMagicArray += ((digest[2] >> 24) & 0xff) % 6;
307 lengthMagicArray += ((digest[2] >> 16) & 0xff) % 6;
308 offsetMagicArray += ((digest[2] >> 8) & 0xff) % 8;
309 offsetMagicArray += ((digest[2] >> 0) & 0xff) % 8;
310 offsetMagicArray += ((digest[3] >> 24) & 0xff) % 8;
311 offsetMagicArray += ((digest[3] >> 16) & 0xff) % 8;
312 offsetMagicArray += ((digest[3] >> 8) & 0xff) % 8;
313 offsetMagicArray += ((digest[3] >> 0) & 0xff) % 8;
314 offsetMagicArray += ((digest[4] >> 24) & 0xff) % 8;
315 offsetMagicArray += ((digest[4] >> 16) & 0xff) % 8;
316 offsetMagicArray += ((digest[4] >> 8) & 0xff) % 8;
317 offsetMagicArray += ((digest[4] >> 0) & 0xff) % 8;
328 for (int i = 0; i < 32; i++) final[i] = 0;
330 final[0] = swap32 (w0[0]);
331 final[1] = swap32 (w0[1]);
332 final[2] = swap32 (w0[2]);
333 final[3] = swap32 (w0[3]);
334 final[4] = swap32 (w1[0]);
335 final[5] = swap32 (w1[1]);
336 final[6] = swap32 (w1[2]);
337 final[7] = swap32 (w1[3]);
339 u32 final_len = pw_len;
345 for (i = 0; i < lengthMagicArray - 4; i += 4)
347 const u32 tmp = GETSHIFTEDINT (theMagicArray, offsetMagicArray + i);
349 SETSHIFTEDINT (final, final_len + i, tmp);
352 const u32 mask = 0xffffffff >> (((i - lengthMagicArray) & 3) * 8);
354 const u32 tmp = GETSHIFTEDINT (theMagicArray, offsetMagicArray + i) & mask;
356 SETSHIFTEDINT (final, final_len + i, tmp);
358 final_len += lengthMagicArray;
362 for (i = 0; i < salt_len + 1; i += 4) // +1 for the 0x80
364 const u32 tmp = salt_buf[i / 4]; // attention, int[] not char[]
366 SETSHIFTEDINT (final, final_len + i, tmp);
369 final_len += salt_len;
376 for (left = final_len, off = 0; left >= 56; left -= 64, off += 16)
378 swap_buffer (&final[off]);
380 sha1_transform (&final[off + 0], &final[off + 4], &final[off + 8], &final[off + 12], digest);
383 swap_buffer (&final[off]);
386 final[off + 15] = final_len * 8;
388 sha1_transform (&final[off + 0], &final[off + 4], &final[off + 8], &final[off + 12], digest);
390 COMPARE_M_SIMD (digest[3], digest[4], digest[2], digest[1]);
394 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_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)
400 const u32 gid = get_global_id (0);
401 const u32 lid = get_local_id (0);
409 salt_buf[0] = salt_bufs[salt_pos].salt_buf[0];
410 salt_buf[1] = salt_bufs[salt_pos].salt_buf[1];
411 salt_buf[2] = salt_bufs[salt_pos].salt_buf[2];
412 salt_buf[3] = salt_bufs[salt_pos].salt_buf[3];
413 salt_buf[4] = salt_bufs[salt_pos].salt_buf[4];
414 salt_buf[5] = salt_bufs[salt_pos].salt_buf[5];
415 salt_buf[6] = salt_bufs[salt_pos].salt_buf[6];
416 salt_buf[7] = salt_bufs[salt_pos].salt_buf[7];
418 const u32 salt_len = salt_bufs[salt_pos].salt_len;
425 s0[0] = swap32 (salt_buf[0]);
426 s0[1] = swap32 (salt_buf[1]);
427 s0[2] = swap32 (salt_buf[2]);
428 s0[3] = swap32 (salt_buf[3]);
429 s1[0] = swap32 (salt_buf[4]);
430 s1[1] = swap32 (salt_buf[5]);
431 s1[2] = swap32 (salt_buf[6]);
432 s1[3] = swap32 (salt_buf[7]);
442 switch_buffer_by_offset_be_S (s0, s1, s2, s3, pw_len);
444 const u32x pw_salt_len = pw_len + salt_len;
450 const u32 search[4] =
452 digests_buf[digests_offset].digest_buf[DGST_R0],
453 digests_buf[digests_offset].digest_buf[DGST_R1],
454 digests_buf[digests_offset].digest_buf[DGST_R2],
455 digests_buf[digests_offset].digest_buf[DGST_R3]
464 for (u32 il_pos = 0; il_pos < il_cnt; il_pos += VECT_SIZE)
466 const u32x w0r = ix_create_bft (bfs_buf, il_pos);
468 const u32x w0lr = w0l | w0r;
478 final[ 0] = w0[0] | s0[0];
479 final[ 1] = w0[1] | s0[1];
480 final[ 2] = w0[2] | s0[2];
481 final[ 3] = w0[3] | s0[3];
482 final[ 4] = w1[0] | s1[0];
483 final[ 5] = w1[1] | s1[1];
484 final[ 6] = w1[2] | s1[2];
485 final[ 7] = w1[3] | s1[3];
486 final[ 8] = w2[0] | s2[0];
487 final[ 9] = w2[1] | s2[1];
488 final[10] = w2[2] | s2[2];
489 final[11] = w2[3] | s2[3];
490 final[12] = w3[0] | s3[0];
491 final[13] = w3[1] | s3[1];
493 final[15] = pw_salt_len * 8;
503 sha1_transform (&final[0], &final[4], &final[8], &final[12], digest);
505 // prepare magic array range
507 u32 lengthMagicArray = 0x20;
508 u32 offsetMagicArray = 0;
510 lengthMagicArray += ((digest[0] >> 24) & 0xff) % 6;
511 lengthMagicArray += ((digest[0] >> 16) & 0xff) % 6;
512 lengthMagicArray += ((digest[0] >> 8) & 0xff) % 6;
513 lengthMagicArray += ((digest[0] >> 0) & 0xff) % 6;
514 lengthMagicArray += ((digest[1] >> 24) & 0xff) % 6;
515 lengthMagicArray += ((digest[1] >> 16) & 0xff) % 6;
516 lengthMagicArray += ((digest[1] >> 8) & 0xff) % 6;
517 lengthMagicArray += ((digest[1] >> 0) & 0xff) % 6;
518 lengthMagicArray += ((digest[2] >> 24) & 0xff) % 6;
519 lengthMagicArray += ((digest[2] >> 16) & 0xff) % 6;
520 offsetMagicArray += ((digest[2] >> 8) & 0xff) % 8;
521 offsetMagicArray += ((digest[2] >> 0) & 0xff) % 8;
522 offsetMagicArray += ((digest[3] >> 24) & 0xff) % 8;
523 offsetMagicArray += ((digest[3] >> 16) & 0xff) % 8;
524 offsetMagicArray += ((digest[3] >> 8) & 0xff) % 8;
525 offsetMagicArray += ((digest[3] >> 0) & 0xff) % 8;
526 offsetMagicArray += ((digest[4] >> 24) & 0xff) % 8;
527 offsetMagicArray += ((digest[4] >> 16) & 0xff) % 8;
528 offsetMagicArray += ((digest[4] >> 8) & 0xff) % 8;
529 offsetMagicArray += ((digest[4] >> 0) & 0xff) % 8;
540 for (int i = 0; i < 32; i++) final[i] = 0;
542 final[0] = swap32 (w0[0]);
543 final[1] = swap32 (w0[1]);
544 final[2] = swap32 (w0[2]);
545 final[3] = swap32 (w0[3]);
546 final[4] = swap32 (w1[0]);
547 final[5] = swap32 (w1[1]);
548 final[6] = swap32 (w1[2]);
549 final[7] = swap32 (w1[3]);
551 u32 final_len = pw_len;
557 for (i = 0; i < lengthMagicArray - 4; i += 4)
559 const u32 tmp = GETSHIFTEDINT (theMagicArray, offsetMagicArray + i);
561 SETSHIFTEDINT (final, final_len + i, tmp);
564 const u32 mask = 0xffffffff >> (((i - lengthMagicArray) & 3) * 8);
566 const u32 tmp = GETSHIFTEDINT (theMagicArray, offsetMagicArray + i) & mask;
568 SETSHIFTEDINT (final, final_len + i, tmp);
570 final_len += lengthMagicArray;
574 for (i = 0; i < salt_len + 1; i += 4) // +1 for the 0x80
576 const u32 tmp = salt_buf[i / 4]; // attention, int[] not char[]
578 SETSHIFTEDINT (final, final_len + i, tmp);
581 final_len += salt_len;
588 for (left = final_len, off = 0; left >= 56; left -= 64, off += 16)
590 swap_buffer (&final[off]);
592 sha1_transform (&final[off + 0], &final[off + 4], &final[off + 8], &final[off + 12], digest);
595 swap_buffer (&final[off]);
598 final[off + 15] = final_len * 8;
600 sha1_transform (&final[off + 0], &final[off + 4], &final[off + 8], &final[off + 12], digest);
602 COMPARE_S_SIMD (digest[3], digest[4], digest[2], digest[1]);
606 __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)
612 const u32 gid = get_global_id (0);
614 if (gid >= gid_max) return;
620 //const u32 lid = get_local_id (0);
624 w0[0] = pws[gid].i[ 0];
625 w0[1] = pws[gid].i[ 1];
626 w0[2] = pws[gid].i[ 2];
627 w0[3] = pws[gid].i[ 3];
650 const u32 pw_len = pws[gid].pw_len;
656 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_scryptV_buf, bitmap_mask, bitmap_shift1, bitmap_shift2, salt_pos, loop_pos, loop_cnt, il_cnt, digests_cnt, digests_offset);
659 __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)
665 const u32 gid = get_global_id (0);
667 if (gid >= gid_max) return;
673 //const u32 lid = get_local_id (0);
677 w0[0] = pws[gid].i[ 0];
678 w0[1] = pws[gid].i[ 1];
679 w0[2] = pws[gid].i[ 2];
680 w0[3] = pws[gid].i[ 3];
684 w1[0] = pws[gid].i[ 4];
685 w1[1] = pws[gid].i[ 5];
686 w1[2] = pws[gid].i[ 6];
687 w1[3] = pws[gid].i[ 7];
703 const u32 pw_len = pws[gid].pw_len;
709 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_scryptV_buf, bitmap_mask, bitmap_shift1, bitmap_shift2, salt_pos, loop_pos, loop_cnt, il_cnt, digests_cnt, digests_offset);
712 __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)
716 __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)
722 const u32 gid = get_global_id (0);
724 if (gid >= gid_max) return;
730 //const u32 lid = get_local_id (0);
734 w0[0] = pws[gid].i[ 0];
735 w0[1] = pws[gid].i[ 1];
736 w0[2] = pws[gid].i[ 2];
737 w0[3] = pws[gid].i[ 3];
760 const u32 pw_len = pws[gid].pw_len;
766 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_scryptV_buf, bitmap_mask, bitmap_shift1, bitmap_shift2, salt_pos, loop_pos, loop_cnt, il_cnt, digests_cnt, digests_offset);
769 __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)
775 const u32 gid = get_global_id (0);
777 if (gid >= gid_max) return;
783 //const u32 lid = get_local_id (0);
787 w0[0] = pws[gid].i[ 0];
788 w0[1] = pws[gid].i[ 1];
789 w0[2] = pws[gid].i[ 2];
790 w0[3] = pws[gid].i[ 3];
794 w1[0] = pws[gid].i[ 4];
795 w1[1] = pws[gid].i[ 5];
796 w1[2] = pws[gid].i[ 6];
797 w1[3] = pws[gid].i[ 7];
813 const u32 pw_len = pws[gid].pw_len;
819 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_scryptV_buf, bitmap_mask, bitmap_shift1, bitmap_shift2, salt_pos, loop_pos, loop_cnt, il_cnt, digests_cnt, digests_offset);
822 __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)