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"
31 #include "include/rp_gpu.h"
35 #define VECT_COMPARE_S "check_single_vect1_comp4.c"
36 #define VECT_COMPARE_M "check_multi_vect1_comp4.c"
40 #define VECT_COMPARE_S "check_single_vect2_comp4.c"
41 #define VECT_COMPARE_M "check_multi_vect2_comp4.c"
45 #define VECT_COMPARE_S "check_single_vect4_comp4.c"
46 #define VECT_COMPARE_M "check_multi_vect4_comp4.c"
49 static void sha1_transform (const u32x w0[4], const u32x w1[4], const u32x w2[4], const u32x w3[4], u32x digest[5])
77 SHA1_STEP (SHA1_F0o, A, B, C, D, E, w0_t);
78 SHA1_STEP (SHA1_F0o, E, A, B, C, D, w1_t);
79 SHA1_STEP (SHA1_F0o, D, E, A, B, C, w2_t);
80 SHA1_STEP (SHA1_F0o, C, D, E, A, B, w3_t);
81 SHA1_STEP (SHA1_F0o, B, C, D, E, A, w4_t);
82 SHA1_STEP (SHA1_F0o, A, B, C, D, E, w5_t);
83 SHA1_STEP (SHA1_F0o, E, A, B, C, D, w6_t);
84 SHA1_STEP (SHA1_F0o, D, E, A, B, C, w7_t);
85 SHA1_STEP (SHA1_F0o, C, D, E, A, B, w8_t);
86 SHA1_STEP (SHA1_F0o, B, C, D, E, A, w9_t);
87 SHA1_STEP (SHA1_F0o, A, B, C, D, E, wa_t);
88 SHA1_STEP (SHA1_F0o, E, A, B, C, D, wb_t);
89 SHA1_STEP (SHA1_F0o, D, E, A, B, C, wc_t);
90 SHA1_STEP (SHA1_F0o, C, D, E, A, B, wd_t);
91 SHA1_STEP (SHA1_F0o, B, C, D, E, A, we_t);
92 SHA1_STEP (SHA1_F0o, A, B, C, D, E, wf_t);
93 w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F0o, E, A, B, C, D, w0_t);
94 w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F0o, D, E, A, B, C, w1_t);
95 w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F0o, C, D, E, A, B, w2_t);
96 w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F0o, B, C, D, E, A, w3_t);
101 w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w4_t);
102 w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w5_t);
103 w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w6_t);
104 w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w7_t);
105 w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w8_t);
106 w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w9_t);
107 wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wa_t);
108 wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, wb_t);
109 wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, wc_t);
110 wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, wd_t);
111 we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, we_t);
112 wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wf_t);
113 w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w0_t);
114 w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w1_t);
115 w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w2_t);
116 w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w3_t);
117 w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w4_t);
118 w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w5_t);
119 w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w6_t);
120 w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w7_t);
125 w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, w8_t);
126 w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, w9_t);
127 wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, wa_t);
128 wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, wb_t);
129 wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, wc_t);
130 wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, wd_t);
131 we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, we_t);
132 wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, wf_t);
133 w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, w0_t);
134 w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, w1_t);
135 w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, w2_t);
136 w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, w3_t);
137 w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, w4_t);
138 w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, w5_t);
139 w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, w6_t);
140 w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, w7_t);
141 w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, w8_t);
142 w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, w9_t);
143 wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, wa_t);
144 wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, wb_t);
149 wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, wc_t);
150 wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wd_t);
151 we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, we_t);
152 wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, wf_t);
153 w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w0_t);
154 w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w1_t);
155 w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w2_t);
156 w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w3_t);
157 w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w4_t);
158 w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w5_t);
159 w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w6_t);
160 w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w7_t);
161 w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w8_t);
162 w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w9_t);
163 wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, wa_t);
164 wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, wb_t);
165 wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wc_t);
166 wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, wd_t);
167 we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, we_t);
168 wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, wf_t);
177 static void hmac_sha1_pad (u32x w0[4], u32x w1[4], u32x w2[4], u32x w3[4], u32x ipad[5], u32x opad[5])
179 w0[0] = w0[0] ^ 0x36363636;
180 w0[1] = w0[1] ^ 0x36363636;
181 w0[2] = w0[2] ^ 0x36363636;
182 w0[3] = w0[3] ^ 0x36363636;
183 w1[0] = w1[0] ^ 0x36363636;
184 w1[1] = w1[1] ^ 0x36363636;
185 w1[2] = w1[2] ^ 0x36363636;
186 w1[3] = w1[3] ^ 0x36363636;
187 w2[0] = w2[0] ^ 0x36363636;
188 w2[1] = w2[1] ^ 0x36363636;
189 w2[2] = w2[2] ^ 0x36363636;
190 w2[3] = w2[3] ^ 0x36363636;
191 w3[0] = w3[0] ^ 0x36363636;
192 w3[1] = w3[1] ^ 0x36363636;
193 w3[2] = w3[2] ^ 0x36363636;
194 w3[3] = w3[3] ^ 0x36363636;
202 sha1_transform (w0, w1, w2, w3, ipad);
204 w0[0] = w0[0] ^ 0x6a6a6a6a;
205 w0[1] = w0[1] ^ 0x6a6a6a6a;
206 w0[2] = w0[2] ^ 0x6a6a6a6a;
207 w0[3] = w0[3] ^ 0x6a6a6a6a;
208 w1[0] = w1[0] ^ 0x6a6a6a6a;
209 w1[1] = w1[1] ^ 0x6a6a6a6a;
210 w1[2] = w1[2] ^ 0x6a6a6a6a;
211 w1[3] = w1[3] ^ 0x6a6a6a6a;
212 w2[0] = w2[0] ^ 0x6a6a6a6a;
213 w2[1] = w2[1] ^ 0x6a6a6a6a;
214 w2[2] = w2[2] ^ 0x6a6a6a6a;
215 w2[3] = w2[3] ^ 0x6a6a6a6a;
216 w3[0] = w3[0] ^ 0x6a6a6a6a;
217 w3[1] = w3[1] ^ 0x6a6a6a6a;
218 w3[2] = w3[2] ^ 0x6a6a6a6a;
219 w3[3] = w3[3] ^ 0x6a6a6a6a;
227 sha1_transform (w0, w1, w2, w3, opad);
230 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])
238 sha1_transform (w0, w1, w2, w3, digest);
255 w3[3] = (64 + 20) * 8;
263 sha1_transform (w0, w1, w2, w3, digest);
266 __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 rules_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
272 const u32 lid = get_local_id (0);
278 const u32 gid = get_global_id (0);
282 pw_buf0[0] = pws[gid].i[ 0];
283 pw_buf0[1] = pws[gid].i[ 1];
284 pw_buf0[2] = pws[gid].i[ 2];
285 pw_buf0[3] = pws[gid].i[ 3];
289 pw_buf1[0] = pws[gid].i[ 4];
290 pw_buf1[1] = pws[gid].i[ 5];
291 pw_buf1[2] = pws[gid].i[ 6];
292 pw_buf1[3] = pws[gid].i[ 7];
294 const u32 pw_len = pws[gid].pw_len;
300 const u32 nr_len = ikepsk_bufs[salt_pos].nr_len;
301 const u32 msg_len = ikepsk_bufs[salt_pos].msg_len;
305 salt_buf0[0] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 0]);
306 salt_buf0[1] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 1]);
307 salt_buf0[2] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 2]);
308 salt_buf0[3] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 3]);
312 salt_buf1[0] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 4]);
313 salt_buf1[1] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 5]);
314 salt_buf1[2] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 6]);
315 salt_buf1[3] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 7]);
319 salt_buf2[0] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 8]);
320 salt_buf2[1] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 9]);
321 salt_buf2[2] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[10]);
322 salt_buf2[3] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[11]);
326 salt_buf3[0] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[12]);
327 salt_buf3[1] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[13]);
331 __local u32 s_msg_buf[128];
333 const u32 lid2 = lid * 2;
335 s_msg_buf[lid2 + 0] = swap_workaround (ikepsk_bufs[salt_pos].msg_buf[lid2 + 0]);
336 s_msg_buf[lid2 + 1] = swap_workaround (ikepsk_bufs[salt_pos].msg_buf[lid2 + 1]);
338 barrier (CLK_LOCAL_MEM_FENCE);
340 if (gid >= gid_max) return;
346 for (u32 il_pos = 0; il_pos < rules_cnt; il_pos++)
376 const u32 out_len = apply_rules (rules_buf[il_pos].cmds, w0, w1, pw_len);
384 w0_t[0] = swap_workaround (w0[0]);
385 w0_t[1] = swap_workaround (w0[1]);
386 w0_t[2] = swap_workaround (w0[2]);
387 w0_t[3] = swap_workaround (w0[3]);
391 w1_t[0] = swap_workaround (w1[0]);
392 w1_t[1] = swap_workaround (w1[1]);
393 w1_t[2] = swap_workaround (w1[2]);
394 w1_t[3] = swap_workaround (w1[3]);
413 hmac_sha1_pad (w0_t, w1_t, w2_t, w3_t, ipad, opad);
415 w0_t[0] = salt_buf0[0];
416 w0_t[1] = salt_buf0[1];
417 w0_t[2] = salt_buf0[2];
418 w0_t[3] = salt_buf0[3];
419 w1_t[0] = salt_buf1[0];
420 w1_t[1] = salt_buf1[1];
421 w1_t[2] = salt_buf1[2];
422 w1_t[3] = salt_buf1[3];
423 w2_t[0] = salt_buf2[0];
424 w2_t[1] = salt_buf2[1];
425 w2_t[2] = salt_buf2[2];
426 w2_t[3] = salt_buf2[3];
427 w3_t[0] = salt_buf3[0];
428 w3_t[1] = salt_buf3[1];
430 w3_t[3] = (64 + nr_len) * 8;
434 hmac_sha1_run (w0_t, w1_t, w2_t, w3_t, ipad, opad, digest);
453 hmac_sha1_pad (w0_t, w1_t, w2_t, w3_t, ipad, opad);
458 for (left = ikepsk_bufs[salt_pos].msg_len, off = 0; left >= 56; left -= 64, off += 16)
460 w0_t[0] = s_msg_buf[off + 0];
461 w0_t[1] = s_msg_buf[off + 1];
462 w0_t[2] = s_msg_buf[off + 2];
463 w0_t[3] = s_msg_buf[off + 3];
464 w1_t[0] = s_msg_buf[off + 4];
465 w1_t[1] = s_msg_buf[off + 5];
466 w1_t[2] = s_msg_buf[off + 6];
467 w1_t[3] = s_msg_buf[off + 7];
468 w2_t[0] = s_msg_buf[off + 8];
469 w2_t[1] = s_msg_buf[off + 9];
470 w2_t[2] = s_msg_buf[off + 10];
471 w2_t[3] = s_msg_buf[off + 11];
472 w3_t[0] = s_msg_buf[off + 12];
473 w3_t[1] = s_msg_buf[off + 13];
474 w3_t[2] = s_msg_buf[off + 14];
475 w3_t[3] = s_msg_buf[off + 15];
477 sha1_transform (w0_t, w1_t, w2_t, w3_t, ipad);
480 w0_t[0] = s_msg_buf[off + 0];
481 w0_t[1] = s_msg_buf[off + 1];
482 w0_t[2] = s_msg_buf[off + 2];
483 w0_t[3] = s_msg_buf[off + 3];
484 w1_t[0] = s_msg_buf[off + 4];
485 w1_t[1] = s_msg_buf[off + 5];
486 w1_t[2] = s_msg_buf[off + 6];
487 w1_t[3] = s_msg_buf[off + 7];
488 w2_t[0] = s_msg_buf[off + 8];
489 w2_t[1] = s_msg_buf[off + 9];
490 w2_t[2] = s_msg_buf[off + 10];
491 w2_t[3] = s_msg_buf[off + 11];
492 w3_t[0] = s_msg_buf[off + 12];
493 w3_t[1] = s_msg_buf[off + 13];
495 w3_t[3] = (64 + msg_len) * 8;
497 hmac_sha1_run (w0_t, w1_t, w2_t, w3_t, ipad, opad, digest);
499 const u32x r0 = digest[3];
500 const u32x r1 = digest[4];
501 const u32x r2 = digest[2];
502 const u32x r3 = digest[1];
504 #include VECT_COMPARE_M
508 __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 combs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
512 __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 combs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
516 __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 rules_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
522 const u32 lid = get_local_id (0);
528 const u32 gid = get_global_id (0);
532 pw_buf0[0] = pws[gid].i[ 0];
533 pw_buf0[1] = pws[gid].i[ 1];
534 pw_buf0[2] = pws[gid].i[ 2];
535 pw_buf0[3] = pws[gid].i[ 3];
539 pw_buf1[0] = pws[gid].i[ 4];
540 pw_buf1[1] = pws[gid].i[ 5];
541 pw_buf1[2] = pws[gid].i[ 6];
542 pw_buf1[3] = pws[gid].i[ 7];
544 const u32 pw_len = pws[gid].pw_len;
550 const u32 nr_len = ikepsk_bufs[salt_pos].nr_len;
551 const u32 msg_len = ikepsk_bufs[salt_pos].msg_len;
555 salt_buf0[0] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 0]);
556 salt_buf0[1] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 1]);
557 salt_buf0[2] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 2]);
558 salt_buf0[3] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 3]);
562 salt_buf1[0] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 4]);
563 salt_buf1[1] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 5]);
564 salt_buf1[2] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 6]);
565 salt_buf1[3] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 7]);
569 salt_buf2[0] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 8]);
570 salt_buf2[1] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[ 9]);
571 salt_buf2[2] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[10]);
572 salt_buf2[3] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[11]);
576 salt_buf3[0] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[12]);
577 salt_buf3[1] = swap_workaround (ikepsk_bufs[salt_pos].nr_buf[13]);
581 __local u32 s_msg_buf[128];
583 const u32 lid2 = lid * 2;
585 s_msg_buf[lid2 + 0] = swap_workaround (ikepsk_bufs[salt_pos].msg_buf[lid2 + 0]);
586 s_msg_buf[lid2 + 1] = swap_workaround (ikepsk_bufs[salt_pos].msg_buf[lid2 + 1]);
588 barrier (CLK_LOCAL_MEM_FENCE);
590 if (gid >= gid_max) return;
596 const u32 search[4] =
598 digests_buf[digests_offset].digest_buf[DGST_R0],
599 digests_buf[digests_offset].digest_buf[DGST_R1],
600 digests_buf[digests_offset].digest_buf[DGST_R2],
601 digests_buf[digests_offset].digest_buf[DGST_R3]
608 for (u32 il_pos = 0; il_pos < rules_cnt; il_pos++)
638 const u32 out_len = apply_rules (rules_buf[il_pos].cmds, w0, w1, pw_len);
646 w0_t[0] = swap_workaround (w0[0]);
647 w0_t[1] = swap_workaround (w0[1]);
648 w0_t[2] = swap_workaround (w0[2]);
649 w0_t[3] = swap_workaround (w0[3]);
653 w1_t[0] = swap_workaround (w1[0]);
654 w1_t[1] = swap_workaround (w1[1]);
655 w1_t[2] = swap_workaround (w1[2]);
656 w1_t[3] = swap_workaround (w1[3]);
675 hmac_sha1_pad (w0_t, w1_t, w2_t, w3_t, ipad, opad);
677 w0_t[0] = salt_buf0[0];
678 w0_t[1] = salt_buf0[1];
679 w0_t[2] = salt_buf0[2];
680 w0_t[3] = salt_buf0[3];
681 w1_t[0] = salt_buf1[0];
682 w1_t[1] = salt_buf1[1];
683 w1_t[2] = salt_buf1[2];
684 w1_t[3] = salt_buf1[3];
685 w2_t[0] = salt_buf2[0];
686 w2_t[1] = salt_buf2[1];
687 w2_t[2] = salt_buf2[2];
688 w2_t[3] = salt_buf2[3];
689 w3_t[0] = salt_buf3[0];
690 w3_t[1] = salt_buf3[1];
692 w3_t[3] = (64 + nr_len) * 8;
696 hmac_sha1_run (w0_t, w1_t, w2_t, w3_t, ipad, opad, digest);
715 hmac_sha1_pad (w0_t, w1_t, w2_t, w3_t, ipad, opad);
720 for (left = ikepsk_bufs[salt_pos].msg_len, off = 0; left >= 56; left -= 64, off += 16)
722 w0_t[0] = s_msg_buf[off + 0];
723 w0_t[1] = s_msg_buf[off + 1];
724 w0_t[2] = s_msg_buf[off + 2];
725 w0_t[3] = s_msg_buf[off + 3];
726 w1_t[0] = s_msg_buf[off + 4];
727 w1_t[1] = s_msg_buf[off + 5];
728 w1_t[2] = s_msg_buf[off + 6];
729 w1_t[3] = s_msg_buf[off + 7];
730 w2_t[0] = s_msg_buf[off + 8];
731 w2_t[1] = s_msg_buf[off + 9];
732 w2_t[2] = s_msg_buf[off + 10];
733 w2_t[3] = s_msg_buf[off + 11];
734 w3_t[0] = s_msg_buf[off + 12];
735 w3_t[1] = s_msg_buf[off + 13];
736 w3_t[2] = s_msg_buf[off + 14];
737 w3_t[3] = s_msg_buf[off + 15];
739 sha1_transform (w0_t, w1_t, w2_t, w3_t, ipad);
742 w0_t[0] = s_msg_buf[off + 0];
743 w0_t[1] = s_msg_buf[off + 1];
744 w0_t[2] = s_msg_buf[off + 2];
745 w0_t[3] = s_msg_buf[off + 3];
746 w1_t[0] = s_msg_buf[off + 4];
747 w1_t[1] = s_msg_buf[off + 5];
748 w1_t[2] = s_msg_buf[off + 6];
749 w1_t[3] = s_msg_buf[off + 7];
750 w2_t[0] = s_msg_buf[off + 8];
751 w2_t[1] = s_msg_buf[off + 9];
752 w2_t[2] = s_msg_buf[off + 10];
753 w2_t[3] = s_msg_buf[off + 11];
754 w3_t[0] = s_msg_buf[off + 12];
755 w3_t[1] = s_msg_buf[off + 13];
757 w3_t[3] = (64 + msg_len) * 8;
759 hmac_sha1_run (w0_t, w1_t, w2_t, w3_t, ipad, opad, digest);
761 const u32x r0 = digest[3];
762 const u32x r1 = digest[4];
763 const u32x r2 = digest[2];
764 const u32x r3 = digest[1];
766 #include VECT_COMPARE_S
770 __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 combs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
774 __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 combs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)