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"
27 #include "include/rp_gpu.h"
31 #define VECT_COMPARE_S "check_single_vect1_comp4.c"
32 #define VECT_COMPARE_M "check_multi_vect1_comp4.c"
36 #define VECT_COMPARE_S "check_single_vect2_comp4.c"
37 #define VECT_COMPARE_M "check_multi_vect2_comp4.c"
41 #define VECT_COMPARE_S "check_single_vect4_comp4.c"
42 #define VECT_COMPARE_M "check_multi_vect4_comp4.c"
45 __device__ static void sha1_transform (const u32x w0[4], const u32x w1[4], const u32x w2[4], const u32x w3[4], u32x digest[5])
73 SHA1_STEP (SHA1_F0o, A, B, C, D, E, w0_t);
74 SHA1_STEP (SHA1_F0o, E, A, B, C, D, w1_t);
75 SHA1_STEP (SHA1_F0o, D, E, A, B, C, w2_t);
76 SHA1_STEP (SHA1_F0o, C, D, E, A, B, w3_t);
77 SHA1_STEP (SHA1_F0o, B, C, D, E, A, w4_t);
78 SHA1_STEP (SHA1_F0o, A, B, C, D, E, w5_t);
79 SHA1_STEP (SHA1_F0o, E, A, B, C, D, w6_t);
80 SHA1_STEP (SHA1_F0o, D, E, A, B, C, w7_t);
81 SHA1_STEP (SHA1_F0o, C, D, E, A, B, w8_t);
82 SHA1_STEP (SHA1_F0o, B, C, D, E, A, w9_t);
83 SHA1_STEP (SHA1_F0o, A, B, C, D, E, wa_t);
84 SHA1_STEP (SHA1_F0o, E, A, B, C, D, wb_t);
85 SHA1_STEP (SHA1_F0o, D, E, A, B, C, wc_t);
86 SHA1_STEP (SHA1_F0o, C, D, E, A, B, wd_t);
87 SHA1_STEP (SHA1_F0o, B, C, D, E, A, we_t);
88 SHA1_STEP (SHA1_F0o, A, B, C, D, E, wf_t);
89 w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F0o, E, A, B, C, D, w0_t);
90 w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F0o, D, E, A, B, C, w1_t);
91 w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F0o, C, D, E, A, B, w2_t);
92 w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F0o, B, C, D, E, A, w3_t);
97 w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w4_t);
98 w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w5_t);
99 w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w6_t);
100 w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w7_t);
101 w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w8_t);
102 w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w9_t);
103 wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wa_t);
104 wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, wb_t);
105 wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, wc_t);
106 wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, wd_t);
107 we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, we_t);
108 wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wf_t);
109 w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w0_t);
110 w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w1_t);
111 w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w2_t);
112 w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w3_t);
113 w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w4_t);
114 w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w5_t);
115 w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w6_t);
116 w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w7_t);
121 w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, w8_t);
122 w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, w9_t);
123 wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, wa_t);
124 wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, wb_t);
125 wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, wc_t);
126 wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, wd_t);
127 we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, we_t);
128 wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, wf_t);
129 w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, w0_t);
130 w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, w1_t);
131 w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, w2_t);
132 w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, w3_t);
133 w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, w4_t);
134 w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, w5_t);
135 w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, w6_t);
136 w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, w7_t);
137 w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, w8_t);
138 w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, w9_t);
139 wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, wa_t);
140 wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, wb_t);
145 wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, wc_t);
146 wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wd_t);
147 we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, we_t);
148 wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, wf_t);
149 w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w0_t);
150 w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w1_t);
151 w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w2_t);
152 w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w3_t);
153 w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w4_t);
154 w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w5_t);
155 w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w6_t);
156 w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w7_t);
157 w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w8_t);
158 w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w9_t);
159 wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, wa_t);
160 wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, wb_t);
161 wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wc_t);
162 wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, wd_t);
163 we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, we_t);
164 wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, wf_t);
173 __device__ static void hmac_sha1_pad (u32x w0[4], u32x w1[4], u32x w2[4], u32x w3[4], u32x ipad[5], u32x opad[5])
175 w0[0] = w0[0] ^ 0x36363636;
176 w0[1] = w0[1] ^ 0x36363636;
177 w0[2] = w0[2] ^ 0x36363636;
178 w0[3] = w0[3] ^ 0x36363636;
179 w1[0] = w1[0] ^ 0x36363636;
180 w1[1] = w1[1] ^ 0x36363636;
181 w1[2] = w1[2] ^ 0x36363636;
182 w1[3] = w1[3] ^ 0x36363636;
183 w2[0] = w2[0] ^ 0x36363636;
184 w2[1] = w2[1] ^ 0x36363636;
185 w2[2] = w2[2] ^ 0x36363636;
186 w2[3] = w2[3] ^ 0x36363636;
187 w3[0] = w3[0] ^ 0x36363636;
188 w3[1] = w3[1] ^ 0x36363636;
189 w3[2] = w3[2] ^ 0x36363636;
190 w3[3] = w3[3] ^ 0x36363636;
198 sha1_transform (w0, w1, w2, w3, ipad);
200 w0[0] = w0[0] ^ 0x6a6a6a6a;
201 w0[1] = w0[1] ^ 0x6a6a6a6a;
202 w0[2] = w0[2] ^ 0x6a6a6a6a;
203 w0[3] = w0[3] ^ 0x6a6a6a6a;
204 w1[0] = w1[0] ^ 0x6a6a6a6a;
205 w1[1] = w1[1] ^ 0x6a6a6a6a;
206 w1[2] = w1[2] ^ 0x6a6a6a6a;
207 w1[3] = w1[3] ^ 0x6a6a6a6a;
208 w2[0] = w2[0] ^ 0x6a6a6a6a;
209 w2[1] = w2[1] ^ 0x6a6a6a6a;
210 w2[2] = w2[2] ^ 0x6a6a6a6a;
211 w2[3] = w2[3] ^ 0x6a6a6a6a;
212 w3[0] = w3[0] ^ 0x6a6a6a6a;
213 w3[1] = w3[1] ^ 0x6a6a6a6a;
214 w3[2] = w3[2] ^ 0x6a6a6a6a;
215 w3[3] = w3[3] ^ 0x6a6a6a6a;
223 sha1_transform (w0, w1, w2, w3, opad);
226 __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])
234 sha1_transform (w0, w1, w2, w3, digest);
251 w3[3] = (64 + 20) * 8;
259 sha1_transform (w0, w1, w2, w3, digest);
262 __device__ __constant__ gpu_rule_t c_rules[1024];
264 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 rules_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
270 const u32 lid = threadIdx.x;
276 const u32 gid = (blockIdx.x * blockDim.x) + threadIdx.x;
280 pw_buf0[0] = pws[gid].i[ 0];
281 pw_buf0[1] = pws[gid].i[ 1];
282 pw_buf0[2] = pws[gid].i[ 2];
283 pw_buf0[3] = pws[gid].i[ 3];
287 pw_buf1[0] = pws[gid].i[ 4];
288 pw_buf1[1] = pws[gid].i[ 5];
289 pw_buf1[2] = pws[gid].i[ 6];
290 pw_buf1[3] = pws[gid].i[ 7];
292 const u32 pw_len = pws[gid].pw_len;
298 const u32 nr_len = ikepsk_bufs[salt_pos].nr_len;
299 const u32 msg_len = ikepsk_bufs[salt_pos].msg_len;
303 salt_buf0[0] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 0]);
304 salt_buf0[1] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 1]);
305 salt_buf0[2] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 2]);
306 salt_buf0[3] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 3]);
310 salt_buf1[0] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 4]);
311 salt_buf1[1] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 5]);
312 salt_buf1[2] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 6]);
313 salt_buf1[3] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 7]);
317 salt_buf2[0] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 8]);
318 salt_buf2[1] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 9]);
319 salt_buf2[2] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[10]);
320 salt_buf2[3] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[11]);
324 salt_buf3[0] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[12]);
325 salt_buf3[1] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[13]);
333 __shared__ u32 s_msg_buf[128];
337 s_msg_buf[lid] = swap_workaround (ikepsk_bufs[salt_pos].msg_buf[lid]);
342 if (gid >= gid_max) return;
348 for (u32 il_pos = 0; il_pos < rules_cnt; il_pos++)
378 const u32 out_len = apply_rules (c_rules[il_pos].cmds, w0, w1, pw_len);
386 w0_t[0] = swap_workaround (w0[0]);
387 w0_t[1] = swap_workaround (w0[1]);
388 w0_t[2] = swap_workaround (w0[2]);
389 w0_t[3] = swap_workaround (w0[3]);
393 w1_t[0] = swap_workaround (w1[0]);
394 w1_t[1] = swap_workaround (w1[1]);
395 w1_t[2] = swap_workaround (w1[2]);
396 w1_t[3] = swap_workaround (w1[3]);
415 hmac_sha1_pad (w0_t, w1_t, w2_t, w3_t, ipad, opad);
417 w0_t[0] = salt_buf0[0];
418 w0_t[1] = salt_buf0[1];
419 w0_t[2] = salt_buf0[2];
420 w0_t[3] = salt_buf0[3];
421 w1_t[0] = salt_buf1[0];
422 w1_t[1] = salt_buf1[1];
423 w1_t[2] = salt_buf1[2];
424 w1_t[3] = salt_buf1[3];
425 w2_t[0] = salt_buf2[0];
426 w2_t[1] = salt_buf2[1];
427 w2_t[2] = salt_buf2[2];
428 w2_t[3] = salt_buf2[3];
429 w3_t[0] = salt_buf3[0];
430 w3_t[1] = salt_buf3[1];
432 w3_t[3] = (64 + nr_len) * 8;
436 hmac_sha1_run (w0_t, w1_t, w2_t, w3_t, ipad, opad, digest);
455 hmac_sha1_pad (w0_t, w1_t, w2_t, w3_t, ipad, opad);
460 for (left = ikepsk_bufs[salt_pos].msg_len, off = 0; left >= 56; left -= 64, off += 16)
462 w0_t[0] = s_msg_buf[off + 0];
463 w0_t[1] = s_msg_buf[off + 1];
464 w0_t[2] = s_msg_buf[off + 2];
465 w0_t[3] = s_msg_buf[off + 3];
466 w1_t[0] = s_msg_buf[off + 4];
467 w1_t[1] = s_msg_buf[off + 5];
468 w1_t[2] = s_msg_buf[off + 6];
469 w1_t[3] = s_msg_buf[off + 7];
470 w2_t[0] = s_msg_buf[off + 8];
471 w2_t[1] = s_msg_buf[off + 9];
472 w2_t[2] = s_msg_buf[off + 10];
473 w2_t[3] = s_msg_buf[off + 11];
474 w3_t[0] = s_msg_buf[off + 12];
475 w3_t[1] = s_msg_buf[off + 13];
476 w3_t[2] = s_msg_buf[off + 14];
477 w3_t[3] = s_msg_buf[off + 15];
479 sha1_transform (w0_t, w1_t, w2_t, w3_t, ipad);
482 w0_t[0] = s_msg_buf[off + 0];
483 w0_t[1] = s_msg_buf[off + 1];
484 w0_t[2] = s_msg_buf[off + 2];
485 w0_t[3] = s_msg_buf[off + 3];
486 w1_t[0] = s_msg_buf[off + 4];
487 w1_t[1] = s_msg_buf[off + 5];
488 w1_t[2] = s_msg_buf[off + 6];
489 w1_t[3] = s_msg_buf[off + 7];
490 w2_t[0] = s_msg_buf[off + 8];
491 w2_t[1] = s_msg_buf[off + 9];
492 w2_t[2] = s_msg_buf[off + 10];
493 w2_t[3] = s_msg_buf[off + 11];
494 w3_t[0] = s_msg_buf[off + 12];
495 w3_t[1] = s_msg_buf[off + 13];
497 w3_t[3] = (64 + msg_len) * 8;
499 hmac_sha1_run (w0_t, w1_t, w2_t, w3_t, ipad, opad, digest);
501 const u32x r0 = digest[3];
502 const u32x r1 = digest[4];
503 const u32x r2 = digest[2];
504 const u32x r3 = digest[1];
506 #include VECT_COMPARE_M
510 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 rules_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
514 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 rules_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
518 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 rules_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
524 const u32 lid = threadIdx.x;
530 const u32 gid = (blockIdx.x * blockDim.x) + threadIdx.x;
534 pw_buf0[0] = pws[gid].i[ 0];
535 pw_buf0[1] = pws[gid].i[ 1];
536 pw_buf0[2] = pws[gid].i[ 2];
537 pw_buf0[3] = pws[gid].i[ 3];
541 pw_buf1[0] = pws[gid].i[ 4];
542 pw_buf1[1] = pws[gid].i[ 5];
543 pw_buf1[2] = pws[gid].i[ 6];
544 pw_buf1[3] = pws[gid].i[ 7];
546 const u32 pw_len = pws[gid].pw_len;
552 const u32 nr_len = ikepsk_bufs[salt_pos].nr_len;
553 const u32 msg_len = ikepsk_bufs[salt_pos].msg_len;
557 salt_buf0[0] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 0]);
558 salt_buf0[1] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 1]);
559 salt_buf0[2] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 2]);
560 salt_buf0[3] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 3]);
564 salt_buf1[0] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 4]);
565 salt_buf1[1] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 5]);
566 salt_buf1[2] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 6]);
567 salt_buf1[3] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 7]);
571 salt_buf2[0] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 8]);
572 salt_buf2[1] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 9]);
573 salt_buf2[2] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[10]);
574 salt_buf2[3] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[11]);
578 salt_buf3[0] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[12]);
579 salt_buf3[1] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[13]);
587 __shared__ u32 s_msg_buf[128];
591 s_msg_buf[lid] = swap_workaround (ikepsk_bufs[salt_pos].msg_buf[lid]);
596 if (gid >= gid_max) return;
602 const u32 search[4] =
604 digests_buf[digests_offset].digest_buf[DGST_R0],
605 digests_buf[digests_offset].digest_buf[DGST_R1],
606 digests_buf[digests_offset].digest_buf[DGST_R2],
607 digests_buf[digests_offset].digest_buf[DGST_R3]
614 for (u32 il_pos = 0; il_pos < rules_cnt; il_pos++)
644 const u32 out_len = apply_rules (c_rules[il_pos].cmds, w0, w1, pw_len);
652 w0_t[0] = swap_workaround (w0[0]);
653 w0_t[1] = swap_workaround (w0[1]);
654 w0_t[2] = swap_workaround (w0[2]);
655 w0_t[3] = swap_workaround (w0[3]);
659 w1_t[0] = swap_workaround (w1[0]);
660 w1_t[1] = swap_workaround (w1[1]);
661 w1_t[2] = swap_workaround (w1[2]);
662 w1_t[3] = swap_workaround (w1[3]);
681 hmac_sha1_pad (w0_t, w1_t, w2_t, w3_t, ipad, opad);
683 w0_t[0] = salt_buf0[0];
684 w0_t[1] = salt_buf0[1];
685 w0_t[2] = salt_buf0[2];
686 w0_t[3] = salt_buf0[3];
687 w1_t[0] = salt_buf1[0];
688 w1_t[1] = salt_buf1[1];
689 w1_t[2] = salt_buf1[2];
690 w1_t[3] = salt_buf1[3];
691 w2_t[0] = salt_buf2[0];
692 w2_t[1] = salt_buf2[1];
693 w2_t[2] = salt_buf2[2];
694 w2_t[3] = salt_buf2[3];
695 w3_t[0] = salt_buf3[0];
696 w3_t[1] = salt_buf3[1];
698 w3_t[3] = (64 + nr_len) * 8;
702 hmac_sha1_run (w0_t, w1_t, w2_t, w3_t, ipad, opad, digest);
721 hmac_sha1_pad (w0_t, w1_t, w2_t, w3_t, ipad, opad);
726 for (left = ikepsk_bufs[salt_pos].msg_len, off = 0; left >= 56; left -= 64, off += 16)
728 w0_t[0] = s_msg_buf[off + 0];
729 w0_t[1] = s_msg_buf[off + 1];
730 w0_t[2] = s_msg_buf[off + 2];
731 w0_t[3] = s_msg_buf[off + 3];
732 w1_t[0] = s_msg_buf[off + 4];
733 w1_t[1] = s_msg_buf[off + 5];
734 w1_t[2] = s_msg_buf[off + 6];
735 w1_t[3] = s_msg_buf[off + 7];
736 w2_t[0] = s_msg_buf[off + 8];
737 w2_t[1] = s_msg_buf[off + 9];
738 w2_t[2] = s_msg_buf[off + 10];
739 w2_t[3] = s_msg_buf[off + 11];
740 w3_t[0] = s_msg_buf[off + 12];
741 w3_t[1] = s_msg_buf[off + 13];
742 w3_t[2] = s_msg_buf[off + 14];
743 w3_t[3] = s_msg_buf[off + 15];
745 sha1_transform (w0_t, w1_t, w2_t, w3_t, ipad);
748 w0_t[0] = s_msg_buf[off + 0];
749 w0_t[1] = s_msg_buf[off + 1];
750 w0_t[2] = s_msg_buf[off + 2];
751 w0_t[3] = s_msg_buf[off + 3];
752 w1_t[0] = s_msg_buf[off + 4];
753 w1_t[1] = s_msg_buf[off + 5];
754 w1_t[2] = s_msg_buf[off + 6];
755 w1_t[3] = s_msg_buf[off + 7];
756 w2_t[0] = s_msg_buf[off + 8];
757 w2_t[1] = s_msg_buf[off + 9];
758 w2_t[2] = s_msg_buf[off + 10];
759 w2_t[3] = s_msg_buf[off + 11];
760 w3_t[0] = s_msg_buf[off + 12];
761 w3_t[1] = s_msg_buf[off + 13];
763 w3_t[3] = (64 + msg_len) * 8;
765 hmac_sha1_run (w0_t, w1_t, w2_t, w3_t, ipad, opad, digest);
767 const u32x r0 = digest[3];
768 const u32x r1 = digest[4];
769 const u32x r2 = digest[2];
770 const u32x r3 = digest[1];
772 #include VECT_COMPARE_S
776 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 rules_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
780 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 rules_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)