2 * Author......: Jens Steube <jens.steube@gmail.com>
8 #include "include/constants.h"
9 #include "include/kernel_vendor.h"
28 #include "include/kernel_functions.c"
29 #include "types_amd.c"
30 #include "common_amd.c"
33 #define VECT_COMPARE_S "check_single_vect1_comp4.c"
34 #define VECT_COMPARE_M "check_multi_vect1_comp4.c"
38 #define VECT_COMPARE_S "check_single_vect2_comp4.c"
39 #define VECT_COMPARE_M "check_multi_vect2_comp4.c"
43 #define VECT_COMPARE_S "check_single_vect4_comp4.c"
44 #define VECT_COMPARE_M "check_multi_vect4_comp4.c"
47 static void sha1_transform (const u32x w0[4], const u32x w1[4], const u32x w2[4], const u32x w3[4], u32x digest[5])
75 SHA1_STEP (SHA1_F0o, A, B, C, D, E, w0_t);
76 SHA1_STEP (SHA1_F0o, E, A, B, C, D, w1_t);
77 SHA1_STEP (SHA1_F0o, D, E, A, B, C, w2_t);
78 SHA1_STEP (SHA1_F0o, C, D, E, A, B, w3_t);
79 SHA1_STEP (SHA1_F0o, B, C, D, E, A, w4_t);
80 SHA1_STEP (SHA1_F0o, A, B, C, D, E, w5_t);
81 SHA1_STEP (SHA1_F0o, E, A, B, C, D, w6_t);
82 SHA1_STEP (SHA1_F0o, D, E, A, B, C, w7_t);
83 SHA1_STEP (SHA1_F0o, C, D, E, A, B, w8_t);
84 SHA1_STEP (SHA1_F0o, B, C, D, E, A, w9_t);
85 SHA1_STEP (SHA1_F0o, A, B, C, D, E, wa_t);
86 SHA1_STEP (SHA1_F0o, E, A, B, C, D, wb_t);
87 SHA1_STEP (SHA1_F0o, D, E, A, B, C, wc_t);
88 SHA1_STEP (SHA1_F0o, C, D, E, A, B, wd_t);
89 SHA1_STEP (SHA1_F0o, B, C, D, E, A, we_t);
90 SHA1_STEP (SHA1_F0o, A, B, C, D, E, wf_t);
91 w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F0o, E, A, B, C, D, w0_t);
92 w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F0o, D, E, A, B, C, w1_t);
93 w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F0o, C, D, E, A, B, w2_t);
94 w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F0o, B, C, D, E, A, w3_t);
99 w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w4_t);
100 w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w5_t);
101 w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w6_t);
102 w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w7_t);
103 w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w8_t);
104 w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w9_t);
105 wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wa_t);
106 wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, wb_t);
107 wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, wc_t);
108 wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, wd_t);
109 we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, we_t);
110 wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wf_t);
111 w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w0_t);
112 w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w1_t);
113 w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w2_t);
114 w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w3_t);
115 w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w4_t);
116 w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w5_t);
117 w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w6_t);
118 w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w7_t);
123 w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, w8_t);
124 w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, w9_t);
125 wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, wa_t);
126 wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, wb_t);
127 wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, wc_t);
128 wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, wd_t);
129 we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, we_t);
130 wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, wf_t);
131 w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, w0_t);
132 w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, w1_t);
133 w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, w2_t);
134 w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, w3_t);
135 w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, w4_t);
136 w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, w5_t);
137 w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, w6_t);
138 w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, w7_t);
139 w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, w8_t);
140 w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, w9_t);
141 wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, wa_t);
142 wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, wb_t);
147 wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, wc_t);
148 wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wd_t);
149 we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, we_t);
150 wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, wf_t);
151 w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w0_t);
152 w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w1_t);
153 w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w2_t);
154 w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w3_t);
155 w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w4_t);
156 w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w5_t);
157 w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w6_t);
158 w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w7_t);
159 w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w8_t);
160 w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w9_t);
161 wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, wa_t);
162 wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, wb_t);
163 wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wc_t);
164 wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, wd_t);
165 we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, we_t);
166 wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, wf_t);
175 static void hmac_sha1_pad (u32x w0[4], u32x w1[4], u32x w2[4], u32x w3[4], u32x ipad[5], u32x opad[5])
177 w0[0] = w0[0] ^ 0x36363636;
178 w0[1] = w0[1] ^ 0x36363636;
179 w0[2] = w0[2] ^ 0x36363636;
180 w0[3] = w0[3] ^ 0x36363636;
181 w1[0] = w1[0] ^ 0x36363636;
182 w1[1] = w1[1] ^ 0x36363636;
183 w1[2] = w1[2] ^ 0x36363636;
184 w1[3] = w1[3] ^ 0x36363636;
185 w2[0] = w2[0] ^ 0x36363636;
186 w2[1] = w2[1] ^ 0x36363636;
187 w2[2] = w2[2] ^ 0x36363636;
188 w2[3] = w2[3] ^ 0x36363636;
189 w3[0] = w3[0] ^ 0x36363636;
190 w3[1] = w3[1] ^ 0x36363636;
191 w3[2] = w3[2] ^ 0x36363636;
192 w3[3] = w3[3] ^ 0x36363636;
200 sha1_transform (w0, w1, w2, w3, ipad);
202 w0[0] = w0[0] ^ 0x6a6a6a6a;
203 w0[1] = w0[1] ^ 0x6a6a6a6a;
204 w0[2] = w0[2] ^ 0x6a6a6a6a;
205 w0[3] = w0[3] ^ 0x6a6a6a6a;
206 w1[0] = w1[0] ^ 0x6a6a6a6a;
207 w1[1] = w1[1] ^ 0x6a6a6a6a;
208 w1[2] = w1[2] ^ 0x6a6a6a6a;
209 w1[3] = w1[3] ^ 0x6a6a6a6a;
210 w2[0] = w2[0] ^ 0x6a6a6a6a;
211 w2[1] = w2[1] ^ 0x6a6a6a6a;
212 w2[2] = w2[2] ^ 0x6a6a6a6a;
213 w2[3] = w2[3] ^ 0x6a6a6a6a;
214 w3[0] = w3[0] ^ 0x6a6a6a6a;
215 w3[1] = w3[1] ^ 0x6a6a6a6a;
216 w3[2] = w3[2] ^ 0x6a6a6a6a;
217 w3[3] = w3[3] ^ 0x6a6a6a6a;
225 sha1_transform (w0, w1, w2, w3, opad);
228 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])
236 sha1_transform (w0, w1, w2, w3, digest);
253 w3[3] = (64 + 20) * 8;
261 sha1_transform (w0, w1, w2, w3, digest);
264 static void m05400m (u32x w0[4], u32x w1[4], u32x w2[4], u32x w3[4], const u32 pw_len, __global pw_t *pws, __global gpu_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 ikepsk_t *ikepsk_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 bfs_cnt, const u32 digests_cnt, const u32 digests_offset, __local u32 s_msg_buf[128])
270 const u32 gid = get_global_id (0);
271 const u32 lid = get_local_id (0);
277 const u32 nr_len = ikepsk_bufs[salt_pos].nr_len;
278 const u32 msg_len = ikepsk_bufs[salt_pos].msg_len;
282 salt_buf0[0] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 0]);
283 salt_buf0[1] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 1]);
284 salt_buf0[2] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 2]);
285 salt_buf0[3] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 3]);
289 salt_buf1[0] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 4]);
290 salt_buf1[1] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 5]);
291 salt_buf1[2] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 6]);
292 salt_buf1[3] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 7]);
296 salt_buf2[0] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 8]);
297 salt_buf2[1] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 9]);
298 salt_buf2[2] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[10]);
299 salt_buf2[3] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[11]);
303 salt_buf3[0] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[12]);
304 salt_buf3[1] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[13]);
314 for (u32 il_pos = 0; il_pos < bfs_cnt; il_pos++)
316 const u32 w0r = bfs_buf[il_pos].i;
355 hmac_sha1_pad (w0_t, w1_t, w2_t, w3_t, ipad, opad);
357 w0_t[0] = salt_buf0[0];
358 w0_t[1] = salt_buf0[1];
359 w0_t[2] = salt_buf0[2];
360 w0_t[3] = salt_buf0[3];
361 w1_t[0] = salt_buf1[0];
362 w1_t[1] = salt_buf1[1];
363 w1_t[2] = salt_buf1[2];
364 w1_t[3] = salt_buf1[3];
365 w2_t[0] = salt_buf2[0];
366 w2_t[1] = salt_buf2[1];
367 w2_t[2] = salt_buf2[2];
368 w2_t[3] = salt_buf2[3];
369 w3_t[0] = salt_buf3[0];
370 w3_t[1] = salt_buf3[1];
372 w3_t[3] = (64 + nr_len) * 8;
376 hmac_sha1_run (w0_t, w1_t, w2_t, w3_t, ipad, opad, digest);
395 hmac_sha1_pad (w0_t, w1_t, w2_t, w3_t, ipad, opad);
400 for (left = ikepsk_bufs[salt_pos].msg_len, off = 0; left >= 56; left -= 64, off += 16)
402 w0_t[0] = s_msg_buf[off + 0];
403 w0_t[1] = s_msg_buf[off + 1];
404 w0_t[2] = s_msg_buf[off + 2];
405 w0_t[3] = s_msg_buf[off + 3];
406 w1_t[0] = s_msg_buf[off + 4];
407 w1_t[1] = s_msg_buf[off + 5];
408 w1_t[2] = s_msg_buf[off + 6];
409 w1_t[3] = s_msg_buf[off + 7];
410 w2_t[0] = s_msg_buf[off + 8];
411 w2_t[1] = s_msg_buf[off + 9];
412 w2_t[2] = s_msg_buf[off + 10];
413 w2_t[3] = s_msg_buf[off + 11];
414 w3_t[0] = s_msg_buf[off + 12];
415 w3_t[1] = s_msg_buf[off + 13];
416 w3_t[2] = s_msg_buf[off + 14];
417 w3_t[3] = s_msg_buf[off + 15];
419 sha1_transform (w0_t, w1_t, w2_t, w3_t, ipad);
422 w0_t[0] = s_msg_buf[off + 0];
423 w0_t[1] = s_msg_buf[off + 1];
424 w0_t[2] = s_msg_buf[off + 2];
425 w0_t[3] = s_msg_buf[off + 3];
426 w1_t[0] = s_msg_buf[off + 4];
427 w1_t[1] = s_msg_buf[off + 5];
428 w1_t[2] = s_msg_buf[off + 6];
429 w1_t[3] = s_msg_buf[off + 7];
430 w2_t[0] = s_msg_buf[off + 8];
431 w2_t[1] = s_msg_buf[off + 9];
432 w2_t[2] = s_msg_buf[off + 10];
433 w2_t[3] = s_msg_buf[off + 11];
434 w3_t[0] = s_msg_buf[off + 12];
435 w3_t[1] = s_msg_buf[off + 13];
437 w3_t[3] = (64 + msg_len) * 8;
439 hmac_sha1_run (w0_t, w1_t, w2_t, w3_t, ipad, opad, digest);
441 const u32x r0 = digest[3];
442 const u32x r1 = digest[4];
443 const u32x r2 = digest[2];
444 const u32x r3 = digest[1];
446 #include VECT_COMPARE_M
450 static void m05400s (u32x w0[4], u32x w1[4], u32x w2[4], u32x w3[4], const u32 pw_len, __global pw_t *pws, __global gpu_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 ikepsk_t *ikepsk_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 bfs_cnt, const u32 digests_cnt, const u32 digests_offset, __local u32 s_msg_buf[128])
456 const u32 gid = get_global_id (0);
457 const u32 lid = get_local_id (0);
463 const u32 nr_len = ikepsk_bufs[salt_pos].nr_len;
464 const u32 msg_len = ikepsk_bufs[salt_pos].msg_len;
468 salt_buf0[0] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 0]);
469 salt_buf0[1] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 1]);
470 salt_buf0[2] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 2]);
471 salt_buf0[3] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 3]);
475 salt_buf1[0] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 4]);
476 salt_buf1[1] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 5]);
477 salt_buf1[2] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 6]);
478 salt_buf1[3] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 7]);
482 salt_buf2[0] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 8]);
483 salt_buf2[1] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 9]);
484 salt_buf2[2] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[10]);
485 salt_buf2[3] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[11]);
489 salt_buf3[0] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[12]);
490 salt_buf3[1] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[13]);
498 const u32 search[4] =
500 digests_buf[digests_offset].digest_buf[DGST_R0],
501 digests_buf[digests_offset].digest_buf[DGST_R1],
502 digests_buf[digests_offset].digest_buf[DGST_R2],
503 digests_buf[digests_offset].digest_buf[DGST_R3]
512 for (u32 il_pos = 0; il_pos < bfs_cnt; il_pos++)
514 const u32 w0r = bfs_buf[il_pos].i;
553 hmac_sha1_pad (w0_t, w1_t, w2_t, w3_t, ipad, opad);
555 w0_t[0] = salt_buf0[0];
556 w0_t[1] = salt_buf0[1];
557 w0_t[2] = salt_buf0[2];
558 w0_t[3] = salt_buf0[3];
559 w1_t[0] = salt_buf1[0];
560 w1_t[1] = salt_buf1[1];
561 w1_t[2] = salt_buf1[2];
562 w1_t[3] = salt_buf1[3];
563 w2_t[0] = salt_buf2[0];
564 w2_t[1] = salt_buf2[1];
565 w2_t[2] = salt_buf2[2];
566 w2_t[3] = salt_buf2[3];
567 w3_t[0] = salt_buf3[0];
568 w3_t[1] = salt_buf3[1];
570 w3_t[3] = (64 + nr_len) * 8;
574 hmac_sha1_run (w0_t, w1_t, w2_t, w3_t, ipad, opad, digest);
593 hmac_sha1_pad (w0_t, w1_t, w2_t, w3_t, ipad, opad);
598 for (left = ikepsk_bufs[salt_pos].msg_len, off = 0; left >= 56; left -= 64, off += 16)
600 w0_t[0] = s_msg_buf[off + 0];
601 w0_t[1] = s_msg_buf[off + 1];
602 w0_t[2] = s_msg_buf[off + 2];
603 w0_t[3] = s_msg_buf[off + 3];
604 w1_t[0] = s_msg_buf[off + 4];
605 w1_t[1] = s_msg_buf[off + 5];
606 w1_t[2] = s_msg_buf[off + 6];
607 w1_t[3] = s_msg_buf[off + 7];
608 w2_t[0] = s_msg_buf[off + 8];
609 w2_t[1] = s_msg_buf[off + 9];
610 w2_t[2] = s_msg_buf[off + 10];
611 w2_t[3] = s_msg_buf[off + 11];
612 w3_t[0] = s_msg_buf[off + 12];
613 w3_t[1] = s_msg_buf[off + 13];
614 w3_t[2] = s_msg_buf[off + 14];
615 w3_t[3] = s_msg_buf[off + 15];
617 sha1_transform (w0_t, w1_t, w2_t, w3_t, ipad);
620 w0_t[0] = s_msg_buf[off + 0];
621 w0_t[1] = s_msg_buf[off + 1];
622 w0_t[2] = s_msg_buf[off + 2];
623 w0_t[3] = s_msg_buf[off + 3];
624 w1_t[0] = s_msg_buf[off + 4];
625 w1_t[1] = s_msg_buf[off + 5];
626 w1_t[2] = s_msg_buf[off + 6];
627 w1_t[3] = s_msg_buf[off + 7];
628 w2_t[0] = s_msg_buf[off + 8];
629 w2_t[1] = s_msg_buf[off + 9];
630 w2_t[2] = s_msg_buf[off + 10];
631 w2_t[3] = s_msg_buf[off + 11];
632 w3_t[0] = s_msg_buf[off + 12];
633 w3_t[1] = s_msg_buf[off + 13];
635 w3_t[3] = (64 + msg_len) * 8;
637 hmac_sha1_run (w0_t, w1_t, w2_t, w3_t, ipad, opad, digest);
639 const u32x r0 = digest[3];
640 const u32x r1 = digest[4];
641 const u32x r2 = digest[2];
642 const u32x r3 = digest[1];
644 #include VECT_COMPARE_S
648 __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m05400_m04 (__global pw_t *pws, __global gpu_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 ikepsk_t *ikepsk_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 bfs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
654 const u32 lid = get_local_id (0);
660 const u32 gid = get_global_id (0);
664 w0[0] = pws[gid].i[ 0];
665 w0[1] = pws[gid].i[ 1];
666 w0[2] = pws[gid].i[ 2];
667 w0[3] = pws[gid].i[ 3];
690 const u32 pw_len = pws[gid].pw_len;
696 __local u32 s_msg_buf[128];
698 const u32 lid2 = lid * 2;
700 s_msg_buf[lid2 + 0] = swap_workaround (ikepsk_bufs[salt_pos].msg_buf[lid2 + 0]);
701 s_msg_buf[lid2 + 1] = swap_workaround (ikepsk_bufs[salt_pos].msg_buf[lid2 + 1]);
703 barrier (CLK_LOCAL_MEM_FENCE);
705 if (gid >= gid_max) return;
711 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);
714 __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m05400_m08 (__global pw_t *pws, __global gpu_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 ikepsk_t *ikepsk_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 bfs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
720 const u32 lid = get_local_id (0);
726 const u32 gid = get_global_id (0);
730 w0[0] = pws[gid].i[ 0];
731 w0[1] = pws[gid].i[ 1];
732 w0[2] = pws[gid].i[ 2];
733 w0[3] = pws[gid].i[ 3];
737 w1[0] = pws[gid].i[ 4];
738 w1[1] = pws[gid].i[ 5];
739 w1[2] = pws[gid].i[ 6];
740 w1[3] = pws[gid].i[ 7];
756 const u32 pw_len = pws[gid].pw_len;
762 __local u32 s_msg_buf[128];
764 const u32 lid2 = lid * 2;
766 s_msg_buf[lid2 + 0] = swap_workaround (ikepsk_bufs[salt_pos].msg_buf[lid2 + 0]);
767 s_msg_buf[lid2 + 1] = swap_workaround (ikepsk_bufs[salt_pos].msg_buf[lid2 + 1]);
769 barrier (CLK_LOCAL_MEM_FENCE);
771 if (gid >= gid_max) return;
777 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);
780 __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m05400_m16 (__global pw_t *pws, __global gpu_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 ikepsk_t *ikepsk_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 bfs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
786 const u32 lid = get_local_id (0);
792 const u32 gid = get_global_id (0);
796 w0[0] = pws[gid].i[ 0];
797 w0[1] = pws[gid].i[ 1];
798 w0[2] = pws[gid].i[ 2];
799 w0[3] = pws[gid].i[ 3];
803 w1[0] = pws[gid].i[ 4];
804 w1[1] = pws[gid].i[ 5];
805 w1[2] = pws[gid].i[ 6];
806 w1[3] = pws[gid].i[ 7];
810 w2[0] = pws[gid].i[ 8];
811 w2[1] = pws[gid].i[ 9];
812 w2[2] = pws[gid].i[10];
813 w2[3] = pws[gid].i[11];
817 w3[0] = pws[gid].i[12];
818 w3[1] = pws[gid].i[13];
822 const u32 pw_len = pws[gid].pw_len;
828 __local u32 s_msg_buf[128];
830 const u32 lid2 = lid * 2;
832 s_msg_buf[lid2 + 0] = swap_workaround (ikepsk_bufs[salt_pos].msg_buf[lid2 + 0]);
833 s_msg_buf[lid2 + 1] = swap_workaround (ikepsk_bufs[salt_pos].msg_buf[lid2 + 1]);
835 barrier (CLK_LOCAL_MEM_FENCE);
837 if (gid >= gid_max) return;
843 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);
846 __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m05400_s04 (__global pw_t *pws, __global gpu_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 ikepsk_t *ikepsk_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 bfs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
852 const u32 lid = get_local_id (0);
858 const u32 gid = get_global_id (0);
862 w0[0] = pws[gid].i[ 0];
863 w0[1] = pws[gid].i[ 1];
864 w0[2] = pws[gid].i[ 2];
865 w0[3] = pws[gid].i[ 3];
888 const u32 pw_len = pws[gid].pw_len;
894 __local u32 s_msg_buf[128];
896 const u32 lid2 = lid * 2;
898 s_msg_buf[lid2 + 0] = swap_workaround (ikepsk_bufs[salt_pos].msg_buf[lid2 + 0]);
899 s_msg_buf[lid2 + 1] = swap_workaround (ikepsk_bufs[salt_pos].msg_buf[lid2 + 1]);
901 barrier (CLK_LOCAL_MEM_FENCE);
903 if (gid >= gid_max) return;
909 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);
912 __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m05400_s08 (__global pw_t *pws, __global gpu_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 ikepsk_t *ikepsk_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 bfs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
918 const u32 lid = get_local_id (0);
924 const u32 gid = get_global_id (0);
928 w0[0] = pws[gid].i[ 0];
929 w0[1] = pws[gid].i[ 1];
930 w0[2] = pws[gid].i[ 2];
931 w0[3] = pws[gid].i[ 3];
935 w1[0] = pws[gid].i[ 4];
936 w1[1] = pws[gid].i[ 5];
937 w1[2] = pws[gid].i[ 6];
938 w1[3] = pws[gid].i[ 7];
954 const u32 pw_len = pws[gid].pw_len;
960 __local u32 s_msg_buf[128];
962 const u32 lid2 = lid * 2;
964 s_msg_buf[lid2 + 0] = swap_workaround (ikepsk_bufs[salt_pos].msg_buf[lid2 + 0]);
965 s_msg_buf[lid2 + 1] = swap_workaround (ikepsk_bufs[salt_pos].msg_buf[lid2 + 1]);
967 barrier (CLK_LOCAL_MEM_FENCE);
969 if (gid >= gid_max) return;
975 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);
978 __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m05400_s16 (__global pw_t *pws, __global gpu_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 ikepsk_t *ikepsk_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 bfs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
984 const u32 lid = get_local_id (0);
990 const u32 gid = get_global_id (0);
994 w0[0] = pws[gid].i[ 0];
995 w0[1] = pws[gid].i[ 1];
996 w0[2] = pws[gid].i[ 2];
997 w0[3] = pws[gid].i[ 3];
1001 w1[0] = pws[gid].i[ 4];
1002 w1[1] = pws[gid].i[ 5];
1003 w1[2] = pws[gid].i[ 6];
1004 w1[3] = pws[gid].i[ 7];
1008 w2[0] = pws[gid].i[ 8];
1009 w2[1] = pws[gid].i[ 9];
1010 w2[2] = pws[gid].i[10];
1011 w2[3] = pws[gid].i[11];
1015 w3[0] = pws[gid].i[12];
1016 w3[1] = pws[gid].i[13];
1020 const u32 pw_len = pws[gid].pw_len;
1026 __local u32 s_msg_buf[128];
1028 const u32 lid2 = lid * 2;
1030 s_msg_buf[lid2 + 0] = swap_workaround (ikepsk_bufs[salt_pos].msg_buf[lid2 + 0]);
1031 s_msg_buf[lid2 + 1] = swap_workaround (ikepsk_bufs[salt_pos].msg_buf[lid2 + 1]);
1033 barrier (CLK_LOCAL_MEM_FENCE);
1035 if (gid >= gid_max) return;
1041 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);