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"
34 #define VECT_COMPARE_S "check_single_vect2_comp4.c"
35 #define VECT_COMPARE_M "check_multi_vect2_comp4.c"
39 #define VECT_COMPARE_S "check_single_vect4_comp4.c"
40 #define VECT_COMPARE_M "check_multi_vect4_comp4.c"
43 __device__ static void sha1_transform (const u32x w0[4], const u32x w1[4], const u32x w2[4], const u32x w3[4], u32x digest[5])
71 SHA1_STEP (SHA1_F0o, A, B, C, D, E, w0_t);
72 SHA1_STEP (SHA1_F0o, E, A, B, C, D, w1_t);
73 SHA1_STEP (SHA1_F0o, D, E, A, B, C, w2_t);
74 SHA1_STEP (SHA1_F0o, C, D, E, A, B, w3_t);
75 SHA1_STEP (SHA1_F0o, B, C, D, E, A, w4_t);
76 SHA1_STEP (SHA1_F0o, A, B, C, D, E, w5_t);
77 SHA1_STEP (SHA1_F0o, E, A, B, C, D, w6_t);
78 SHA1_STEP (SHA1_F0o, D, E, A, B, C, w7_t);
79 SHA1_STEP (SHA1_F0o, C, D, E, A, B, w8_t);
80 SHA1_STEP (SHA1_F0o, B, C, D, E, A, w9_t);
81 SHA1_STEP (SHA1_F0o, A, B, C, D, E, wa_t);
82 SHA1_STEP (SHA1_F0o, E, A, B, C, D, wb_t);
83 SHA1_STEP (SHA1_F0o, D, E, A, B, C, wc_t);
84 SHA1_STEP (SHA1_F0o, C, D, E, A, B, wd_t);
85 SHA1_STEP (SHA1_F0o, B, C, D, E, A, we_t);
86 SHA1_STEP (SHA1_F0o, A, B, C, D, E, wf_t);
87 w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F0o, E, A, B, C, D, w0_t);
88 w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F0o, D, E, A, B, C, w1_t);
89 w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F0o, C, D, E, A, B, w2_t);
90 w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F0o, B, C, D, E, A, w3_t);
95 w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w4_t);
96 w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w5_t);
97 w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w6_t);
98 w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w7_t);
99 w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w8_t);
100 w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w9_t);
101 wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wa_t);
102 wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, wb_t);
103 wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, wc_t);
104 wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, wd_t);
105 we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, we_t);
106 wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wf_t);
107 w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w0_t);
108 w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w1_t);
109 w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w2_t);
110 w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w3_t);
111 w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w4_t);
112 w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w5_t);
113 w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w6_t);
114 w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w7_t);
119 w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, w8_t);
120 w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, w9_t);
121 wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, wa_t);
122 wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, wb_t);
123 wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, wc_t);
124 wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, wd_t);
125 we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, we_t);
126 wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, wf_t);
127 w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, w0_t);
128 w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, w1_t);
129 w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, w2_t);
130 w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, w3_t);
131 w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, w4_t);
132 w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, w5_t);
133 w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, w6_t);
134 w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, w7_t);
135 w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, w8_t);
136 w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, w9_t);
137 wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, wa_t);
138 wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, wb_t);
143 wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, wc_t);
144 wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wd_t);
145 we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, we_t);
146 wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, wf_t);
147 w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w0_t);
148 w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w1_t);
149 w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w2_t);
150 w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w3_t);
151 w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w4_t);
152 w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w5_t);
153 w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w6_t);
154 w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w7_t);
155 w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w8_t);
156 w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w9_t);
157 wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, wa_t);
158 wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, wb_t);
159 wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wc_t);
160 wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, wd_t);
161 we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, we_t);
162 wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, wf_t);
171 __device__ static void hmac_sha1_pad (u32x w0[4], u32x w1[4], u32x w2[4], u32x w3[4], u32x ipad[5], u32x opad[5])
173 w0[0] = w0[0] ^ 0x36363636;
174 w0[1] = w0[1] ^ 0x36363636;
175 w0[2] = w0[2] ^ 0x36363636;
176 w0[3] = w0[3] ^ 0x36363636;
177 w1[0] = w1[0] ^ 0x36363636;
178 w1[1] = w1[1] ^ 0x36363636;
179 w1[2] = w1[2] ^ 0x36363636;
180 w1[3] = w1[3] ^ 0x36363636;
181 w2[0] = w2[0] ^ 0x36363636;
182 w2[1] = w2[1] ^ 0x36363636;
183 w2[2] = w2[2] ^ 0x36363636;
184 w2[3] = w2[3] ^ 0x36363636;
185 w3[0] = w3[0] ^ 0x36363636;
186 w3[1] = w3[1] ^ 0x36363636;
187 w3[2] = w3[2] ^ 0x36363636;
188 w3[3] = w3[3] ^ 0x36363636;
196 sha1_transform (w0, w1, w2, w3, ipad);
198 w0[0] = w0[0] ^ 0x6a6a6a6a;
199 w0[1] = w0[1] ^ 0x6a6a6a6a;
200 w0[2] = w0[2] ^ 0x6a6a6a6a;
201 w0[3] = w0[3] ^ 0x6a6a6a6a;
202 w1[0] = w1[0] ^ 0x6a6a6a6a;
203 w1[1] = w1[1] ^ 0x6a6a6a6a;
204 w1[2] = w1[2] ^ 0x6a6a6a6a;
205 w1[3] = w1[3] ^ 0x6a6a6a6a;
206 w2[0] = w2[0] ^ 0x6a6a6a6a;
207 w2[1] = w2[1] ^ 0x6a6a6a6a;
208 w2[2] = w2[2] ^ 0x6a6a6a6a;
209 w2[3] = w2[3] ^ 0x6a6a6a6a;
210 w3[0] = w3[0] ^ 0x6a6a6a6a;
211 w3[1] = w3[1] ^ 0x6a6a6a6a;
212 w3[2] = w3[2] ^ 0x6a6a6a6a;
213 w3[3] = w3[3] ^ 0x6a6a6a6a;
221 sha1_transform (w0, w1, w2, w3, opad);
224 __device__ static void hmac_sha1_run (u32x w0[4], u32x w1[4], u32x w2[4], u32x w3[4], u32x ipad[5], u32x opad[5], u32x digest[5])
232 sha1_transform (w0, w1, w2, w3, digest);
249 w3[3] = (64 + 20) * 8;
257 sha1_transform (w0, w1, w2, w3, digest);
260 __device__ __constant__ bf_t c_bfs[1024];
262 __device__ static void m05400m (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 ikepsk_t *ikepsk_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, u32 s_msg_buf[128])
268 const u32 gid = (blockIdx.x * blockDim.x) + threadIdx.x;
269 const u32 lid = threadIdx.x;
275 const u32 nr_len = ikepsk_bufs[salt_pos].nr_len;
276 const u32 msg_len = ikepsk_bufs[salt_pos].msg_len;
280 salt_buf0[0] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 0]);
281 salt_buf0[1] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 1]);
282 salt_buf0[2] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 2]);
283 salt_buf0[3] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 3]);
287 salt_buf1[0] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 4]);
288 salt_buf1[1] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 5]);
289 salt_buf1[2] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 6]);
290 salt_buf1[3] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 7]);
294 salt_buf2[0] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 8]);
295 salt_buf2[1] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 9]);
296 salt_buf2[2] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[10]);
297 salt_buf2[3] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[11]);
301 salt_buf3[0] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[12]);
302 salt_buf3[1] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[13]);
312 for (u32 il_pos = 0; il_pos < bfs_cnt; il_pos++)
314 const u32 w0r = c_bfs[il_pos].i;
353 hmac_sha1_pad (w0_t, w1_t, w2_t, w3_t, ipad, opad);
355 w0_t[0] = salt_buf0[0];
356 w0_t[1] = salt_buf0[1];
357 w0_t[2] = salt_buf0[2];
358 w0_t[3] = salt_buf0[3];
359 w1_t[0] = salt_buf1[0];
360 w1_t[1] = salt_buf1[1];
361 w1_t[2] = salt_buf1[2];
362 w1_t[3] = salt_buf1[3];
363 w2_t[0] = salt_buf2[0];
364 w2_t[1] = salt_buf2[1];
365 w2_t[2] = salt_buf2[2];
366 w2_t[3] = salt_buf2[3];
367 w3_t[0] = salt_buf3[0];
368 w3_t[1] = salt_buf3[1];
370 w3_t[3] = (64 + nr_len) * 8;
374 hmac_sha1_run (w0_t, w1_t, w2_t, w3_t, ipad, opad, digest);
393 hmac_sha1_pad (w0_t, w1_t, w2_t, w3_t, ipad, opad);
398 for (left = ikepsk_bufs[salt_pos].msg_len, off = 0; left >= 56; left -= 64, off += 16)
400 w0_t[0] = s_msg_buf[off + 0];
401 w0_t[1] = s_msg_buf[off + 1];
402 w0_t[2] = s_msg_buf[off + 2];
403 w0_t[3] = s_msg_buf[off + 3];
404 w1_t[0] = s_msg_buf[off + 4];
405 w1_t[1] = s_msg_buf[off + 5];
406 w1_t[2] = s_msg_buf[off + 6];
407 w1_t[3] = s_msg_buf[off + 7];
408 w2_t[0] = s_msg_buf[off + 8];
409 w2_t[1] = s_msg_buf[off + 9];
410 w2_t[2] = s_msg_buf[off + 10];
411 w2_t[3] = s_msg_buf[off + 11];
412 w3_t[0] = s_msg_buf[off + 12];
413 w3_t[1] = s_msg_buf[off + 13];
414 w3_t[2] = s_msg_buf[off + 14];
415 w3_t[3] = s_msg_buf[off + 15];
417 sha1_transform (w0_t, w1_t, w2_t, w3_t, ipad);
420 w0_t[0] = s_msg_buf[off + 0];
421 w0_t[1] = s_msg_buf[off + 1];
422 w0_t[2] = s_msg_buf[off + 2];
423 w0_t[3] = s_msg_buf[off + 3];
424 w1_t[0] = s_msg_buf[off + 4];
425 w1_t[1] = s_msg_buf[off + 5];
426 w1_t[2] = s_msg_buf[off + 6];
427 w1_t[3] = s_msg_buf[off + 7];
428 w2_t[0] = s_msg_buf[off + 8];
429 w2_t[1] = s_msg_buf[off + 9];
430 w2_t[2] = s_msg_buf[off + 10];
431 w2_t[3] = s_msg_buf[off + 11];
432 w3_t[0] = s_msg_buf[off + 12];
433 w3_t[1] = s_msg_buf[off + 13];
435 w3_t[3] = (64 + msg_len) * 8;
437 hmac_sha1_run (w0_t, w1_t, w2_t, w3_t, ipad, opad, digest);
439 const u32x r0 = digest[3];
440 const u32x r1 = digest[4];
441 const u32x r2 = digest[2];
442 const u32x r3 = digest[1];
444 #include VECT_COMPARE_M
448 __device__ static void m05400s (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 ikepsk_t *ikepsk_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, u32 s_msg_buf[128])
454 const u32 gid = (blockIdx.x * blockDim.x) + threadIdx.x;
455 const u32 lid = threadIdx.x;
461 const u32 nr_len = ikepsk_bufs[salt_pos].nr_len;
462 const u32 msg_len = ikepsk_bufs[salt_pos].msg_len;
466 salt_buf0[0] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 0]);
467 salt_buf0[1] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 1]);
468 salt_buf0[2] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 2]);
469 salt_buf0[3] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 3]);
473 salt_buf1[0] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 4]);
474 salt_buf1[1] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 5]);
475 salt_buf1[2] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 6]);
476 salt_buf1[3] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 7]);
480 salt_buf2[0] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 8]);
481 salt_buf2[1] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 9]);
482 salt_buf2[2] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[10]);
483 salt_buf2[3] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[11]);
487 salt_buf3[0] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[12]);
488 salt_buf3[1] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[13]);
496 const u32 search[4] =
498 digests_buf[digests_offset].digest_buf[DGST_R0],
499 digests_buf[digests_offset].digest_buf[DGST_R1],
500 digests_buf[digests_offset].digest_buf[DGST_R2],
501 digests_buf[digests_offset].digest_buf[DGST_R3]
510 for (u32 il_pos = 0; il_pos < bfs_cnt; il_pos++)
512 const u32 w0r = c_bfs[il_pos].i;
551 hmac_sha1_pad (w0_t, w1_t, w2_t, w3_t, ipad, opad);
553 w0_t[0] = salt_buf0[0];
554 w0_t[1] = salt_buf0[1];
555 w0_t[2] = salt_buf0[2];
556 w0_t[3] = salt_buf0[3];
557 w1_t[0] = salt_buf1[0];
558 w1_t[1] = salt_buf1[1];
559 w1_t[2] = salt_buf1[2];
560 w1_t[3] = salt_buf1[3];
561 w2_t[0] = salt_buf2[0];
562 w2_t[1] = salt_buf2[1];
563 w2_t[2] = salt_buf2[2];
564 w2_t[3] = salt_buf2[3];
565 w3_t[0] = salt_buf3[0];
566 w3_t[1] = salt_buf3[1];
568 w3_t[3] = (64 + nr_len) * 8;
572 hmac_sha1_run (w0_t, w1_t, w2_t, w3_t, ipad, opad, digest);
591 hmac_sha1_pad (w0_t, w1_t, w2_t, w3_t, ipad, opad);
596 for (left = ikepsk_bufs[salt_pos].msg_len, off = 0; left >= 56; left -= 64, off += 16)
598 w0_t[0] = s_msg_buf[off + 0];
599 w0_t[1] = s_msg_buf[off + 1];
600 w0_t[2] = s_msg_buf[off + 2];
601 w0_t[3] = s_msg_buf[off + 3];
602 w1_t[0] = s_msg_buf[off + 4];
603 w1_t[1] = s_msg_buf[off + 5];
604 w1_t[2] = s_msg_buf[off + 6];
605 w1_t[3] = s_msg_buf[off + 7];
606 w2_t[0] = s_msg_buf[off + 8];
607 w2_t[1] = s_msg_buf[off + 9];
608 w2_t[2] = s_msg_buf[off + 10];
609 w2_t[3] = s_msg_buf[off + 11];
610 w3_t[0] = s_msg_buf[off + 12];
611 w3_t[1] = s_msg_buf[off + 13];
612 w3_t[2] = s_msg_buf[off + 14];
613 w3_t[3] = s_msg_buf[off + 15];
615 sha1_transform (w0_t, w1_t, w2_t, w3_t, ipad);
618 w0_t[0] = s_msg_buf[off + 0];
619 w0_t[1] = s_msg_buf[off + 1];
620 w0_t[2] = s_msg_buf[off + 2];
621 w0_t[3] = s_msg_buf[off + 3];
622 w1_t[0] = s_msg_buf[off + 4];
623 w1_t[1] = s_msg_buf[off + 5];
624 w1_t[2] = s_msg_buf[off + 6];
625 w1_t[3] = s_msg_buf[off + 7];
626 w2_t[0] = s_msg_buf[off + 8];
627 w2_t[1] = s_msg_buf[off + 9];
628 w2_t[2] = s_msg_buf[off + 10];
629 w2_t[3] = s_msg_buf[off + 11];
630 w3_t[0] = s_msg_buf[off + 12];
631 w3_t[1] = s_msg_buf[off + 13];
633 w3_t[3] = (64 + msg_len) * 8;
635 hmac_sha1_run (w0_t, w1_t, w2_t, w3_t, ipad, opad, digest);
637 const u32x r0 = digest[3];
638 const u32x r1 = digest[4];
639 const u32x r2 = digest[2];
640 const u32x r3 = digest[1];
642 #include VECT_COMPARE_S
646 extern "C" __global__ void __launch_bounds__ (256, 1) m05400_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 ikepsk_t *ikepsk_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)
652 const u32 gid = (blockIdx.x * blockDim.x) + threadIdx.x;
653 const u32 lid = threadIdx.x;
657 w0[0] = pws[gid].i[ 0];
658 w0[1] = pws[gid].i[ 1];
659 w0[2] = pws[gid].i[ 2];
660 w0[3] = pws[gid].i[ 3];
683 const u32 pw_len = pws[gid].pw_len;
689 __shared__ u32 s_msg_buf[128];
693 s_msg_buf[lid] = swap_workaround (ikepsk_bufs[salt_pos].msg_buf[lid]);
698 if (gid >= gid_max) return;
704 m05400m (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, ikepsk_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, s_msg_buf);
707 extern "C" __global__ void __launch_bounds__ (256, 1) m05400_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 ikepsk_t *ikepsk_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)
713 const u32 gid = (blockIdx.x * blockDim.x) + threadIdx.x;
714 const u32 lid = threadIdx.x;
718 w0[0] = pws[gid].i[ 0];
719 w0[1] = pws[gid].i[ 1];
720 w0[2] = pws[gid].i[ 2];
721 w0[3] = pws[gid].i[ 3];
725 w1[0] = pws[gid].i[ 4];
726 w1[1] = pws[gid].i[ 5];
727 w1[2] = pws[gid].i[ 6];
728 w1[3] = pws[gid].i[ 7];
744 const u32 pw_len = pws[gid].pw_len;
750 __shared__ u32 s_msg_buf[128];
754 s_msg_buf[lid] = swap_workaround (ikepsk_bufs[salt_pos].msg_buf[lid]);
759 if (gid >= gid_max) return;
765 m05400m (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, ikepsk_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, s_msg_buf);
768 extern "C" __global__ void __launch_bounds__ (256, 1) m05400_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 ikepsk_t *ikepsk_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)
774 const u32 gid = (blockIdx.x * blockDim.x) + threadIdx.x;
775 const u32 lid = threadIdx.x;
779 w0[0] = pws[gid].i[ 0];
780 w0[1] = pws[gid].i[ 1];
781 w0[2] = pws[gid].i[ 2];
782 w0[3] = pws[gid].i[ 3];
786 w1[0] = pws[gid].i[ 4];
787 w1[1] = pws[gid].i[ 5];
788 w1[2] = pws[gid].i[ 6];
789 w1[3] = pws[gid].i[ 7];
793 w2[0] = pws[gid].i[ 8];
794 w2[1] = pws[gid].i[ 9];
795 w2[2] = pws[gid].i[10];
796 w2[3] = pws[gid].i[11];
800 w3[0] = pws[gid].i[12];
801 w3[1] = pws[gid].i[13];
805 const u32 pw_len = pws[gid].pw_len;
811 __shared__ u32 s_msg_buf[128];
815 s_msg_buf[lid] = swap_workaround (ikepsk_bufs[salt_pos].msg_buf[lid]);
820 if (gid >= gid_max) return;
826 m05400m (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, ikepsk_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, s_msg_buf);
829 extern "C" __global__ void __launch_bounds__ (256, 1) m05400_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 ikepsk_t *ikepsk_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)
835 const u32 gid = (blockIdx.x * blockDim.x) + threadIdx.x;
836 const u32 lid = threadIdx.x;
840 w0[0] = pws[gid].i[ 0];
841 w0[1] = pws[gid].i[ 1];
842 w0[2] = pws[gid].i[ 2];
843 w0[3] = pws[gid].i[ 3];
866 const u32 pw_len = pws[gid].pw_len;
872 __shared__ u32 s_msg_buf[128];
876 s_msg_buf[lid] = swap_workaround (ikepsk_bufs[salt_pos].msg_buf[lid]);
881 if (gid >= gid_max) return;
887 m05400s (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, ikepsk_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, s_msg_buf);
890 extern "C" __global__ void __launch_bounds__ (256, 1) m05400_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 ikepsk_t *ikepsk_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)
896 const u32 gid = (blockIdx.x * blockDim.x) + threadIdx.x;
897 const u32 lid = threadIdx.x;
901 w0[0] = pws[gid].i[ 0];
902 w0[1] = pws[gid].i[ 1];
903 w0[2] = pws[gid].i[ 2];
904 w0[3] = pws[gid].i[ 3];
908 w1[0] = pws[gid].i[ 4];
909 w1[1] = pws[gid].i[ 5];
910 w1[2] = pws[gid].i[ 6];
911 w1[3] = pws[gid].i[ 7];
927 const u32 pw_len = pws[gid].pw_len;
933 __shared__ u32 s_msg_buf[128];
937 s_msg_buf[lid] = swap_workaround (ikepsk_bufs[salt_pos].msg_buf[lid]);
942 if (gid >= gid_max) return;
948 m05400s (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, ikepsk_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, s_msg_buf);
951 extern "C" __global__ void __launch_bounds__ (256, 1) m05400_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 ikepsk_t *ikepsk_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)
957 const u32 gid = (blockIdx.x * blockDim.x) + threadIdx.x;
958 const u32 lid = threadIdx.x;
962 w0[0] = pws[gid].i[ 0];
963 w0[1] = pws[gid].i[ 1];
964 w0[2] = pws[gid].i[ 2];
965 w0[3] = pws[gid].i[ 3];
969 w1[0] = pws[gid].i[ 4];
970 w1[1] = pws[gid].i[ 5];
971 w1[2] = pws[gid].i[ 6];
972 w1[3] = pws[gid].i[ 7];
976 w2[0] = pws[gid].i[ 8];
977 w2[1] = pws[gid].i[ 9];
978 w2[2] = pws[gid].i[10];
979 w2[3] = pws[gid].i[11];
983 w3[0] = pws[gid].i[12];
984 w3[1] = pws[gid].i[13];
988 const u32 pw_len = pws[gid].pw_len;
994 __shared__ u32 s_msg_buf[128];
998 s_msg_buf[lid] = swap_workaround (ikepsk_bufs[salt_pos].msg_buf[lid]);
1003 if (gid >= gid_max) return;
1009 m05400s (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, ikepsk_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, s_msg_buf);