2 * Author......: Jens Steube <jens.steube@gmail.com>
8 #include "include/constants.h"
9 #include "include/kernel_vendor.h"
24 #include "include/kernel_functions.c"
26 #include "common_nv.c"
29 #define VECT_COMPARE_S "check_single_vect1_comp4.c"
30 #define VECT_COMPARE_M "check_multi_vect1_comp4.c"
33 __device__ static u32 bytealign (u32 src0, u32 src1, u32 src2)
35 return (u32) (((((u64)src0) << 32) | (u64)src1) >> ((src2 & 3)*8));
38 #if __CUDA_ARCH__ >= 350
39 #define GETSHIFTEDINT(a,n) __funnelshift_r ((a)[((n)/4)+0], (a)[((n)/4)+1], (n & 3) * 8)
40 #elif __CUDA_ARCH__ >= 200
41 #define GETSHIFTEDINT(a,n) __byte_perm ((a)[((n)/4)+0], (a)[((n)/4)+1], (0x76543210 >> ((n & 3) * 4)) & 0xffff)
43 #define GETSHIFTEDINT(a,n) bytealign ((a)[((n)/4)+1], (a)[((n)/4)+0], n)
46 #define SETSHIFTEDINT(a,n,v) \
48 const u32 s = ((n) & 3) * 8; \
49 const u64 x = (u64) (v) << s; \
50 (a)[((n)/4)+0] |= x; \
51 (a)[((n)/4)+1] = x >> 32; \
54 __device__ __constant__ u32 theMagicArray[64] =
56 0x1451ac91,0x4354679f,0xe03be724,0xc27b7428,0xeb133386,0x5ccb4f5a,0x37730a08,0x2f1c5d0e,
57 0xe5e68f33,0xddae9bf8,0x8d4bf216,0xdcd4e12c,0x9ddfcbb0,0x176d70d4,0x3f424df9,0x94111b9b,
58 0x9bc15b9f,0x039d0506,0x8a135e9d,0xe86a9a1e,0x17147cd9,0xf62ac758,0x0a6399a1,0xc370fdd7,
59 0x13745ef6,0x040bc903,0x26f79826,0x2593928a,0x230da2b0,0x6d7963ed,0x3cfa3213,0xa39a0235,
60 0x0a8eddb3,0xc351bf24,0x9f55cd7c,0x4c94af37,0x82520829,0x374e3bb2,0x9107179f,0xcdfd3b11,
61 0, 0, 0, 0, 0, 0, 0, 0,
62 0, 0, 0, 0, 0, 0, 0, 0,
63 0, 0, 0, 0, 0, 0, 0, 0
66 __device__ static void swap_buffer (u32x final[16])
68 final[ 0] = swap_workaround (final[ 0]);
69 final[ 1] = swap_workaround (final[ 1]);
70 final[ 2] = swap_workaround (final[ 2]);
71 final[ 3] = swap_workaround (final[ 3]);
72 final[ 4] = swap_workaround (final[ 4]);
73 final[ 5] = swap_workaround (final[ 5]);
74 final[ 6] = swap_workaround (final[ 6]);
75 final[ 7] = swap_workaround (final[ 7]);
76 final[ 8] = swap_workaround (final[ 8]);
77 final[ 9] = swap_workaround (final[ 9]);
78 final[10] = swap_workaround (final[10]);
79 final[11] = swap_workaround (final[11]);
80 final[12] = swap_workaround (final[12]);
81 final[13] = swap_workaround (final[13]);
82 final[14] = swap_workaround (final[14]);
83 final[15] = swap_workaround (final[15]);
86 __device__ static void sha1_transform (const u32x w0[4], const u32x w1[4], const u32x w2[4], const u32x w3[4], u32x digest[5])
114 SHA1_STEP (SHA1_F0o, A, B, C, D, E, w0_t);
115 SHA1_STEP (SHA1_F0o, E, A, B, C, D, w1_t);
116 SHA1_STEP (SHA1_F0o, D, E, A, B, C, w2_t);
117 SHA1_STEP (SHA1_F0o, C, D, E, A, B, w3_t);
118 SHA1_STEP (SHA1_F0o, B, C, D, E, A, w4_t);
119 SHA1_STEP (SHA1_F0o, A, B, C, D, E, w5_t);
120 SHA1_STEP (SHA1_F0o, E, A, B, C, D, w6_t);
121 SHA1_STEP (SHA1_F0o, D, E, A, B, C, w7_t);
122 SHA1_STEP (SHA1_F0o, C, D, E, A, B, w8_t);
123 SHA1_STEP (SHA1_F0o, B, C, D, E, A, w9_t);
124 SHA1_STEP (SHA1_F0o, A, B, C, D, E, wa_t);
125 SHA1_STEP (SHA1_F0o, E, A, B, C, D, wb_t);
126 SHA1_STEP (SHA1_F0o, D, E, A, B, C, wc_t);
127 SHA1_STEP (SHA1_F0o, C, D, E, A, B, wd_t);
128 SHA1_STEP (SHA1_F0o, B, C, D, E, A, we_t);
129 SHA1_STEP (SHA1_F0o, A, B, C, D, E, wf_t);
130 w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F0o, E, A, B, C, D, w0_t);
131 w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F0o, D, E, A, B, C, w1_t);
132 w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F0o, C, D, E, A, B, w2_t);
133 w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F0o, B, C, D, E, A, w3_t);
138 w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w4_t);
139 w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w5_t);
140 w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w6_t);
141 w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w7_t);
142 w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w8_t);
143 w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w9_t);
144 wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wa_t);
145 wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, wb_t);
146 wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, wc_t);
147 wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, wd_t);
148 we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, we_t);
149 wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wf_t);
150 w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w0_t);
151 w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w1_t);
152 w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w2_t);
153 w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w3_t);
154 w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w4_t);
155 w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w5_t);
156 w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w6_t);
157 w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w7_t);
162 w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, w8_t);
163 w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, w9_t);
164 wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, wa_t);
165 wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, wb_t);
166 wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, wc_t);
167 wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, wd_t);
168 we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, we_t);
169 wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, wf_t);
170 w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, w0_t);
171 w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, w1_t);
172 w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, w2_t);
173 w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, w3_t);
174 w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, w4_t);
175 w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, w5_t);
176 w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, w6_t);
177 w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, w7_t);
178 w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, w8_t);
179 w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, w9_t);
180 wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, wa_t);
181 wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, wb_t);
186 wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, wc_t);
187 wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wd_t);
188 we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, we_t);
189 wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, wf_t);
190 w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w0_t);
191 w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w1_t);
192 w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w2_t);
193 w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w3_t);
194 w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w4_t);
195 w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w5_t);
196 w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w6_t);
197 w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w7_t);
198 w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w8_t);
199 w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w9_t);
200 wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, wa_t);
201 wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, wb_t);
202 wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wc_t);
203 wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, wd_t);
204 we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, we_t);
205 wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, wf_t);
214 __device__ __constant__ bf_t c_bfs[1024];
216 __device__ static void m07800m (u32x w0[4], u32x w1[4], u32x w2[4], u32x w3[4], const u32 pw_len, const pw_t *pws, const gpu_rule_t *rules_buf, const comb_t *combs_buf, const bf_t *bfs_buf, const void *tmps, void *hooks, const u32 *bitmaps_buf_s1_a, const u32 *bitmaps_buf_s1_b, const u32 *bitmaps_buf_s1_c, const u32 *bitmaps_buf_s1_d, const u32 *bitmaps_buf_s2_a, const u32 *bitmaps_buf_s2_b, const u32 *bitmaps_buf_s2_c, const u32 *bitmaps_buf_s2_d, plain_t *plains_buf, const digest_t *digests_buf, u32 *hashes_shown, const salt_t *salt_bufs, const void *esalt_bufs, u32 *d_return_buf, 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 bfs_cnt, const u32 digests_cnt, const u32 digests_offset)
222 const u32 gid = (blockIdx.x * blockDim.x) + threadIdx.x;
223 const u32 lid = threadIdx.x;
225 w0[0] = swap_workaround (w0[0]);
226 w0[1] = swap_workaround (w0[1]);
227 w0[2] = swap_workaround (w0[2]);
228 w0[3] = swap_workaround (w0[3]);
230 w1[0] = swap_workaround (w1[0]);
231 w1[1] = swap_workaround (w1[1]);
232 w1[2] = swap_workaround (w1[2]);
233 w1[3] = swap_workaround (w1[3]);
241 salt_buf[0] = salt_bufs[salt_pos].salt_buf[0];
242 salt_buf[1] = salt_bufs[salt_pos].salt_buf[1];
243 salt_buf[2] = salt_bufs[salt_pos].salt_buf[2];
244 salt_buf[3] = salt_bufs[salt_pos].salt_buf[3];
245 salt_buf[4] = salt_bufs[salt_pos].salt_buf[4];
246 salt_buf[5] = salt_bufs[salt_pos].salt_buf[5];
247 salt_buf[6] = salt_bufs[salt_pos].salt_buf[6];
248 salt_buf[7] = salt_bufs[salt_pos].salt_buf[7];
250 const u32 salt_len = salt_bufs[salt_pos].salt_len;
280 switch_buffer_by_offset (s0, s1, s2, s3, pw_len);
282 const u32 pw_salt_len = pw_len + salt_len;
290 for (u32 il_pos = 0; il_pos < bfs_cnt; il_pos++)
292 const u32 w0r = swap_workaround (c_bfs[il_pos].i);
302 final[ 0] = swap_workaround (w0[0] | s0[0]);
303 final[ 1] = swap_workaround (w0[1] | s0[1]);
304 final[ 2] = swap_workaround (w0[2] | s0[2]);
305 final[ 3] = swap_workaround (w0[3] | s0[3]);
306 final[ 4] = swap_workaround (w1[0] | s1[0]);
307 final[ 5] = swap_workaround (w1[1] | s1[1]);
308 final[ 6] = swap_workaround (w1[2] | s1[2]);
309 final[ 7] = swap_workaround (w1[3] | s1[3]);
310 final[ 8] = swap_workaround (w2[0] | s2[0]);
311 final[ 9] = swap_workaround (w2[1] | s2[1]);
312 final[10] = swap_workaround (w2[2] | s2[2]);
313 final[11] = swap_workaround (w2[3] | s2[3]);
314 final[12] = swap_workaround (w3[0] | s3[0]);
315 final[13] = swap_workaround (w3[1] | s3[1]);
317 final[15] = pw_salt_len * 8;
327 sha1_transform (&final[0], &final[4], &final[8], &final[12], digest);
329 // prepare magic array range
331 u32x lengthMagicArray = 0x20;
332 u32x offsetMagicArray = 0;
334 lengthMagicArray += ((digest[0] >> 24) & 0xff) % 6;
335 lengthMagicArray += ((digest[0] >> 16) & 0xff) % 6;
336 lengthMagicArray += ((digest[0] >> 8) & 0xff) % 6;
337 lengthMagicArray += ((digest[0] >> 0) & 0xff) % 6;
338 lengthMagicArray += ((digest[1] >> 24) & 0xff) % 6;
339 lengthMagicArray += ((digest[1] >> 16) & 0xff) % 6;
340 lengthMagicArray += ((digest[1] >> 8) & 0xff) % 6;
341 lengthMagicArray += ((digest[1] >> 0) & 0xff) % 6;
342 lengthMagicArray += ((digest[2] >> 24) & 0xff) % 6;
343 lengthMagicArray += ((digest[2] >> 16) & 0xff) % 6;
344 offsetMagicArray += ((digest[2] >> 8) & 0xff) % 8;
345 offsetMagicArray += ((digest[2] >> 0) & 0xff) % 8;
346 offsetMagicArray += ((digest[3] >> 24) & 0xff) % 8;
347 offsetMagicArray += ((digest[3] >> 16) & 0xff) % 8;
348 offsetMagicArray += ((digest[3] >> 8) & 0xff) % 8;
349 offsetMagicArray += ((digest[3] >> 0) & 0xff) % 8;
350 offsetMagicArray += ((digest[4] >> 24) & 0xff) % 8;
351 offsetMagicArray += ((digest[4] >> 16) & 0xff) % 8;
352 offsetMagicArray += ((digest[4] >> 8) & 0xff) % 8;
353 offsetMagicArray += ((digest[4] >> 0) & 0xff) % 8;
364 for (int i = 0; i < 64; i++) final[i] = 0;
375 u32 final_len = pw_len;
381 for (i = 0; i < lengthMagicArray - 4; i += 4)
383 const u32 tmp = GETSHIFTEDINT (theMagicArray, offsetMagicArray + i);
385 SETSHIFTEDINT (final, final_len + i, tmp);
388 const u32 mask = 0xffffffff >> (((i - lengthMagicArray) & 3) * 8);
390 const u32 tmp = GETSHIFTEDINT (theMagicArray, offsetMagicArray + i) & mask;
392 SETSHIFTEDINT (final, final_len + i, tmp);
394 final_len += lengthMagicArray;
398 for (i = 0; i < salt_len + 1; i += 4) // +1 for the 0x80
400 const u32 tmp = salt_buf[i / 4]; // attention, int[] not char[]
402 SETSHIFTEDINT (final, final_len + i, tmp);
405 final_len += salt_len;
412 for (left = final_len, off = 0; left >= 56; left -= 64, off += 16)
414 swap_buffer (&final[off]);
416 sha1_transform (&final[off + 0], &final[off + 4], &final[off + 8], &final[off + 12], digest);
419 swap_buffer (&final[off]);
422 final[off + 15] = final_len * 8;
424 sha1_transform (&final[off + 0], &final[off + 4], &final[off + 8], &final[off + 12], digest);
426 const u32x r0 = digest[3];
427 const u32x r1 = digest[4];
428 const u32x r2 = digest[2];
429 const u32x r3 = digest[1];
431 #include VECT_COMPARE_M
435 __device__ static void m07800s (u32x w0[4], u32x w1[4], u32x w2[4], u32x w3[4], const u32 pw_len, const pw_t *pws, const gpu_rule_t *rules_buf, const comb_t *combs_buf, const bf_t *bfs_buf, const void *tmps, void *hooks, const u32 *bitmaps_buf_s1_a, const u32 *bitmaps_buf_s1_b, const u32 *bitmaps_buf_s1_c, const u32 *bitmaps_buf_s1_d, const u32 *bitmaps_buf_s2_a, const u32 *bitmaps_buf_s2_b, const u32 *bitmaps_buf_s2_c, const u32 *bitmaps_buf_s2_d, plain_t *plains_buf, const digest_t *digests_buf, u32 *hashes_shown, const salt_t *salt_bufs, const void *esalt_bufs, u32 *d_return_buf, 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 bfs_cnt, const u32 digests_cnt, const u32 digests_offset)
441 const u32 gid = (blockIdx.x * blockDim.x) + threadIdx.x;
442 const u32 lid = threadIdx.x;
444 w0[0] = swap_workaround (w0[0]);
445 w0[1] = swap_workaround (w0[1]);
446 w0[2] = swap_workaround (w0[2]);
447 w0[3] = swap_workaround (w0[3]);
449 w1[0] = swap_workaround (w1[0]);
450 w1[1] = swap_workaround (w1[1]);
451 w1[2] = swap_workaround (w1[2]);
452 w1[3] = swap_workaround (w1[3]);
460 salt_buf[0] = salt_bufs[salt_pos].salt_buf[0];
461 salt_buf[1] = salt_bufs[salt_pos].salt_buf[1];
462 salt_buf[2] = salt_bufs[salt_pos].salt_buf[2];
463 salt_buf[3] = salt_bufs[salt_pos].salt_buf[3];
464 salt_buf[4] = salt_bufs[salt_pos].salt_buf[4];
465 salt_buf[5] = salt_bufs[salt_pos].salt_buf[5];
466 salt_buf[6] = salt_bufs[salt_pos].salt_buf[6];
467 salt_buf[7] = salt_bufs[salt_pos].salt_buf[7];
469 const u32 salt_len = salt_bufs[salt_pos].salt_len;
499 switch_buffer_by_offset (s0, s1, s2, s3, pw_len);
501 const u32 pw_salt_len = pw_len + salt_len;
507 const u32 search[4] =
509 digests_buf[digests_offset].digest_buf[DGST_R0],
510 digests_buf[digests_offset].digest_buf[DGST_R1],
511 digests_buf[digests_offset].digest_buf[DGST_R2],
512 digests_buf[digests_offset].digest_buf[DGST_R3]
521 for (u32 il_pos = 0; il_pos < bfs_cnt; il_pos++)
523 const u32 w0r = swap_workaround (c_bfs[il_pos].i);
533 final[ 0] = swap_workaround (w0[0] | s0[0]);
534 final[ 1] = swap_workaround (w0[1] | s0[1]);
535 final[ 2] = swap_workaround (w0[2] | s0[2]);
536 final[ 3] = swap_workaround (w0[3] | s0[3]);
537 final[ 4] = swap_workaround (w1[0] | s1[0]);
538 final[ 5] = swap_workaround (w1[1] | s1[1]);
539 final[ 6] = swap_workaround (w1[2] | s1[2]);
540 final[ 7] = swap_workaround (w1[3] | s1[3]);
541 final[ 8] = swap_workaround (w2[0] | s2[0]);
542 final[ 9] = swap_workaround (w2[1] | s2[1]);
543 final[10] = swap_workaround (w2[2] | s2[2]);
544 final[11] = swap_workaround (w2[3] | s2[3]);
545 final[12] = swap_workaround (w3[0] | s3[0]);
546 final[13] = swap_workaround (w3[1] | s3[1]);
548 final[15] = pw_salt_len * 8;
558 sha1_transform (&final[0], &final[4], &final[8], &final[12], digest);
560 // prepare magic array range
562 u32x lengthMagicArray = 0x20;
563 u32x offsetMagicArray = 0;
565 lengthMagicArray += ((digest[0] >> 24) & 0xff) % 6;
566 lengthMagicArray += ((digest[0] >> 16) & 0xff) % 6;
567 lengthMagicArray += ((digest[0] >> 8) & 0xff) % 6;
568 lengthMagicArray += ((digest[0] >> 0) & 0xff) % 6;
569 lengthMagicArray += ((digest[1] >> 24) & 0xff) % 6;
570 lengthMagicArray += ((digest[1] >> 16) & 0xff) % 6;
571 lengthMagicArray += ((digest[1] >> 8) & 0xff) % 6;
572 lengthMagicArray += ((digest[1] >> 0) & 0xff) % 6;
573 lengthMagicArray += ((digest[2] >> 24) & 0xff) % 6;
574 lengthMagicArray += ((digest[2] >> 16) & 0xff) % 6;
575 offsetMagicArray += ((digest[2] >> 8) & 0xff) % 8;
576 offsetMagicArray += ((digest[2] >> 0) & 0xff) % 8;
577 offsetMagicArray += ((digest[3] >> 24) & 0xff) % 8;
578 offsetMagicArray += ((digest[3] >> 16) & 0xff) % 8;
579 offsetMagicArray += ((digest[3] >> 8) & 0xff) % 8;
580 offsetMagicArray += ((digest[3] >> 0) & 0xff) % 8;
581 offsetMagicArray += ((digest[4] >> 24) & 0xff) % 8;
582 offsetMagicArray += ((digest[4] >> 16) & 0xff) % 8;
583 offsetMagicArray += ((digest[4] >> 8) & 0xff) % 8;
584 offsetMagicArray += ((digest[4] >> 0) & 0xff) % 8;
595 for (int i = 0; i < 64; i++) final[i] = 0;
606 u32 final_len = pw_len;
612 for (i = 0; i < lengthMagicArray - 4; i += 4)
614 const u32 tmp = GETSHIFTEDINT (theMagicArray, offsetMagicArray + i);
616 SETSHIFTEDINT (final, final_len + i, tmp);
619 const u32 mask = 0xffffffff >> (((i - lengthMagicArray) & 3) * 8);
621 const u32 tmp = GETSHIFTEDINT (theMagicArray, offsetMagicArray + i) & mask;
623 SETSHIFTEDINT (final, final_len + i, tmp);
625 final_len += lengthMagicArray;
629 for (i = 0; i < salt_len + 1; i += 4) // +1 for the 0x80
631 const u32 tmp = salt_buf[i / 4]; // attention, int[] not char[]
633 SETSHIFTEDINT (final, final_len + i, tmp);
636 final_len += salt_len;
643 for (left = final_len, off = 0; left >= 56; left -= 64, off += 16)
645 swap_buffer (&final[off]);
647 sha1_transform (&final[off + 0], &final[off + 4], &final[off + 8], &final[off + 12], digest);
650 swap_buffer (&final[off]);
653 final[off + 15] = final_len * 8;
655 sha1_transform (&final[off + 0], &final[off + 4], &final[off + 8], &final[off + 12], digest);
657 const u32x r0 = digest[3];
658 const u32x r1 = digest[4];
659 const u32x r2 = digest[2];
660 const u32x r3 = digest[1];
662 #include VECT_COMPARE_S
666 extern "C" __global__ void __launch_bounds__ (256, 1) m07800_m04 (const pw_t *pws, const gpu_rule_t *rules_buf, const comb_t *combs_buf, const bf_t *bfs_buf, const void *tmps, void *hooks, const u32 *bitmaps_buf_s1_a, const u32 *bitmaps_buf_s1_b, const u32 *bitmaps_buf_s1_c, const u32 *bitmaps_buf_s1_d, const u32 *bitmaps_buf_s2_a, const u32 *bitmaps_buf_s2_b, const u32 *bitmaps_buf_s2_c, const u32 *bitmaps_buf_s2_d, plain_t *plains_buf, const digest_t *digests_buf, u32 *hashes_shown, const salt_t *salt_bufs, const void *esalt_bufs, u32 *d_return_buf, 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 bfs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
672 const u32 gid = (blockIdx.x * blockDim.x) + threadIdx.x;
674 if (gid >= gid_max) return;
678 w0[0] = pws[gid].i[ 0];
679 w0[1] = pws[gid].i[ 1];
680 w0[2] = pws[gid].i[ 2];
681 w0[3] = pws[gid].i[ 3];
704 const u32 pw_len = pws[gid].pw_len;
710 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, bfs_cnt, digests_cnt, digests_offset);
713 extern "C" __global__ void __launch_bounds__ (256, 1) m07800_m08 (const pw_t *pws, const gpu_rule_t *rules_buf, const comb_t *combs_buf, const bf_t *bfs_buf, const void *tmps, void *hooks, const u32 *bitmaps_buf_s1_a, const u32 *bitmaps_buf_s1_b, const u32 *bitmaps_buf_s1_c, const u32 *bitmaps_buf_s1_d, const u32 *bitmaps_buf_s2_a, const u32 *bitmaps_buf_s2_b, const u32 *bitmaps_buf_s2_c, const u32 *bitmaps_buf_s2_d, plain_t *plains_buf, const digest_t *digests_buf, u32 *hashes_shown, const salt_t *salt_bufs, const void *esalt_bufs, u32 *d_return_buf, 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 bfs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
719 const u32 gid = (blockIdx.x * blockDim.x) + threadIdx.x;
721 if (gid >= gid_max) return;
725 w0[0] = pws[gid].i[ 0];
726 w0[1] = pws[gid].i[ 1];
727 w0[2] = pws[gid].i[ 2];
728 w0[3] = pws[gid].i[ 3];
732 w1[0] = pws[gid].i[ 4];
733 w1[1] = pws[gid].i[ 5];
734 w1[2] = pws[gid].i[ 6];
735 w1[3] = pws[gid].i[ 7];
751 const u32 pw_len = pws[gid].pw_len;
757 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, bfs_cnt, digests_cnt, digests_offset);
760 extern "C" __global__ void __launch_bounds__ (256, 1) m07800_m16 (const pw_t *pws, const gpu_rule_t *rules_buf, const comb_t *combs_buf, const bf_t *bfs_buf, const void *tmps, void *hooks, const u32 *bitmaps_buf_s1_a, const u32 *bitmaps_buf_s1_b, const u32 *bitmaps_buf_s1_c, const u32 *bitmaps_buf_s1_d, const u32 *bitmaps_buf_s2_a, const u32 *bitmaps_buf_s2_b, const u32 *bitmaps_buf_s2_c, const u32 *bitmaps_buf_s2_d, plain_t *plains_buf, const digest_t *digests_buf, u32 *hashes_shown, const salt_t *salt_bufs, const void *esalt_bufs, u32 *d_return_buf, 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 bfs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
764 extern "C" __global__ void __launch_bounds__ (256, 1) m07800_s04 (const pw_t *pws, const gpu_rule_t *rules_buf, const comb_t *combs_buf, const bf_t *bfs_buf, const void *tmps, void *hooks, const u32 *bitmaps_buf_s1_a, const u32 *bitmaps_buf_s1_b, const u32 *bitmaps_buf_s1_c, const u32 *bitmaps_buf_s1_d, const u32 *bitmaps_buf_s2_a, const u32 *bitmaps_buf_s2_b, const u32 *bitmaps_buf_s2_c, const u32 *bitmaps_buf_s2_d, plain_t *plains_buf, const digest_t *digests_buf, u32 *hashes_shown, const salt_t *salt_bufs, const void *esalt_bufs, u32 *d_return_buf, 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 bfs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
770 const u32 gid = (blockIdx.x * blockDim.x) + threadIdx.x;
772 if (gid >= gid_max) return;
776 w0[0] = pws[gid].i[ 0];
777 w0[1] = pws[gid].i[ 1];
778 w0[2] = pws[gid].i[ 2];
779 w0[3] = pws[gid].i[ 3];
802 const u32 pw_len = pws[gid].pw_len;
808 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, bfs_cnt, digests_cnt, digests_offset);
811 extern "C" __global__ void __launch_bounds__ (256, 1) m07800_s08 (const pw_t *pws, const gpu_rule_t *rules_buf, const comb_t *combs_buf, const bf_t *bfs_buf, const void *tmps, void *hooks, const u32 *bitmaps_buf_s1_a, const u32 *bitmaps_buf_s1_b, const u32 *bitmaps_buf_s1_c, const u32 *bitmaps_buf_s1_d, const u32 *bitmaps_buf_s2_a, const u32 *bitmaps_buf_s2_b, const u32 *bitmaps_buf_s2_c, const u32 *bitmaps_buf_s2_d, plain_t *plains_buf, const digest_t *digests_buf, u32 *hashes_shown, const salt_t *salt_bufs, const void *esalt_bufs, u32 *d_return_buf, 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 bfs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
817 const u32 gid = (blockIdx.x * blockDim.x) + threadIdx.x;
819 if (gid >= gid_max) return;
823 w0[0] = pws[gid].i[ 0];
824 w0[1] = pws[gid].i[ 1];
825 w0[2] = pws[gid].i[ 2];
826 w0[3] = pws[gid].i[ 3];
830 w1[0] = pws[gid].i[ 4];
831 w1[1] = pws[gid].i[ 5];
832 w1[2] = pws[gid].i[ 6];
833 w1[3] = pws[gid].i[ 7];
849 const u32 pw_len = pws[gid].pw_len;
855 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, bfs_cnt, digests_cnt, digests_offset);
858 extern "C" __global__ void __launch_bounds__ (256, 1) m07800_s16 (const pw_t *pws, const gpu_rule_t *rules_buf, const comb_t *combs_buf, const bf_t *bfs_buf, const void *tmps, void *hooks, const u32 *bitmaps_buf_s1_a, const u32 *bitmaps_buf_s1_b, const u32 *bitmaps_buf_s1_c, const u32 *bitmaps_buf_s1_d, const u32 *bitmaps_buf_s2_a, const u32 *bitmaps_buf_s2_b, const u32 *bitmaps_buf_s2_c, const u32 *bitmaps_buf_s2_d, plain_t *plains_buf, const digest_t *digests_buf, u32 *hashes_shown, const salt_t *salt_bufs, const void *esalt_bufs, u32 *d_return_buf, 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 bfs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)