2 * Author......: Jens Steube <jens.steube@gmail.com>
8 #include "include/constants.h"
9 #include "include/kernel_vendor.h"
16 #include "include/kernel_functions.c"
17 #include "OpenCL/types_ocl.c"
18 #include "OpenCL/common.c"
20 #define COMPARE_S "OpenCL/check_single_comp4.c"
21 #define COMPARE_M "OpenCL/check_multi_comp4.c"
23 static void sha1_transform (const u32 w0[4], const u32 w1[4], const u32 w2[4], const u32 w3[4], u32 digest[5])
51 SHA1_STEP (SHA1_F0o, A, B, C, D, E, w0_t);
52 SHA1_STEP (SHA1_F0o, E, A, B, C, D, w1_t);
53 SHA1_STEP (SHA1_F0o, D, E, A, B, C, w2_t);
54 SHA1_STEP (SHA1_F0o, C, D, E, A, B, w3_t);
55 SHA1_STEP (SHA1_F0o, B, C, D, E, A, w4_t);
56 SHA1_STEP (SHA1_F0o, A, B, C, D, E, w5_t);
57 SHA1_STEP (SHA1_F0o, E, A, B, C, D, w6_t);
58 SHA1_STEP (SHA1_F0o, D, E, A, B, C, w7_t);
59 SHA1_STEP (SHA1_F0o, C, D, E, A, B, w8_t);
60 SHA1_STEP (SHA1_F0o, B, C, D, E, A, w9_t);
61 SHA1_STEP (SHA1_F0o, A, B, C, D, E, wa_t);
62 SHA1_STEP (SHA1_F0o, E, A, B, C, D, wb_t);
63 SHA1_STEP (SHA1_F0o, D, E, A, B, C, wc_t);
64 SHA1_STEP (SHA1_F0o, C, D, E, A, B, wd_t);
65 SHA1_STEP (SHA1_F0o, B, C, D, E, A, we_t);
66 SHA1_STEP (SHA1_F0o, A, B, C, D, E, wf_t);
67 w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F0o, E, A, B, C, D, w0_t);
68 w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F0o, D, E, A, B, C, w1_t);
69 w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F0o, C, D, E, A, B, w2_t);
70 w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F0o, B, C, D, E, A, w3_t);
75 w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w4_t);
76 w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w5_t);
77 w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w6_t);
78 w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w7_t);
79 w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w8_t);
80 w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w9_t);
81 wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wa_t);
82 wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, wb_t);
83 wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, wc_t);
84 wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, wd_t);
85 we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, we_t);
86 wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wf_t);
87 w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w0_t);
88 w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w1_t);
89 w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w2_t);
90 w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w3_t);
91 w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w4_t);
92 w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w5_t);
93 w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w6_t);
94 w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w7_t);
99 w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, w8_t);
100 w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, w9_t);
101 wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, wa_t);
102 wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, wb_t);
103 wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, wc_t);
104 wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, wd_t);
105 we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, we_t);
106 wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, wf_t);
107 w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, w0_t);
108 w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, w1_t);
109 w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, w2_t);
110 w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, w3_t);
111 w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, w4_t);
112 w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, w5_t);
113 w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, w6_t);
114 w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, w7_t);
115 w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, w8_t);
116 w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, w9_t);
117 wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, wa_t);
118 wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, wb_t);
123 wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, wc_t);
124 wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wd_t);
125 we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, we_t);
126 wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, wf_t);
127 w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w0_t);
128 w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w1_t);
129 w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w2_t);
130 w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w3_t);
131 w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w4_t);
132 w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w5_t);
133 w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w6_t);
134 w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w7_t);
135 w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w8_t);
136 w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w9_t);
137 wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, wa_t);
138 wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, wb_t);
139 wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wc_t);
140 wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, wd_t);
141 we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, we_t);
142 wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, wf_t);
151 static void hmac_sha1_pad (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], u32 ipad[5], u32 opad[5])
153 w0[0] = w0[0] ^ 0x36363636;
154 w0[1] = w0[1] ^ 0x36363636;
155 w0[2] = w0[2] ^ 0x36363636;
156 w0[3] = w0[3] ^ 0x36363636;
157 w1[0] = w1[0] ^ 0x36363636;
158 w1[1] = w1[1] ^ 0x36363636;
159 w1[2] = w1[2] ^ 0x36363636;
160 w1[3] = w1[3] ^ 0x36363636;
161 w2[0] = w2[0] ^ 0x36363636;
162 w2[1] = w2[1] ^ 0x36363636;
163 w2[2] = w2[2] ^ 0x36363636;
164 w2[3] = w2[3] ^ 0x36363636;
165 w3[0] = w3[0] ^ 0x36363636;
166 w3[1] = w3[1] ^ 0x36363636;
167 w3[2] = w3[2] ^ 0x36363636;
168 w3[3] = w3[3] ^ 0x36363636;
176 sha1_transform (w0, w1, w2, w3, ipad);
178 w0[0] = w0[0] ^ 0x6a6a6a6a;
179 w0[1] = w0[1] ^ 0x6a6a6a6a;
180 w0[2] = w0[2] ^ 0x6a6a6a6a;
181 w0[3] = w0[3] ^ 0x6a6a6a6a;
182 w1[0] = w1[0] ^ 0x6a6a6a6a;
183 w1[1] = w1[1] ^ 0x6a6a6a6a;
184 w1[2] = w1[2] ^ 0x6a6a6a6a;
185 w1[3] = w1[3] ^ 0x6a6a6a6a;
186 w2[0] = w2[0] ^ 0x6a6a6a6a;
187 w2[1] = w2[1] ^ 0x6a6a6a6a;
188 w2[2] = w2[2] ^ 0x6a6a6a6a;
189 w2[3] = w2[3] ^ 0x6a6a6a6a;
190 w3[0] = w3[0] ^ 0x6a6a6a6a;
191 w3[1] = w3[1] ^ 0x6a6a6a6a;
192 w3[2] = w3[2] ^ 0x6a6a6a6a;
193 w3[3] = w3[3] ^ 0x6a6a6a6a;
201 sha1_transform (w0, w1, w2, w3, opad);
204 static void hmac_sha1_run (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], u32 ipad[5], u32 opad[5], u32 digest[5])
212 sha1_transform (w0, w1, w2, w3, digest);
229 w3[3] = (64 + 20) * 8;
237 sha1_transform (w0, w1, w2, w3, digest);
240 __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 combs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
246 const u32 lid = get_local_id (0);
252 const u32 gid = get_global_id (0);
256 wordl0[0] = pws[gid].i[ 0];
257 wordl0[1] = pws[gid].i[ 1];
258 wordl0[2] = pws[gid].i[ 2];
259 wordl0[3] = pws[gid].i[ 3];
263 wordl1[0] = pws[gid].i[ 4];
264 wordl1[1] = pws[gid].i[ 5];
265 wordl1[2] = pws[gid].i[ 6];
266 wordl1[3] = pws[gid].i[ 7];
282 const u32 pw_l_len = pws[gid].pw_len;
284 if (combs_mode == COMBINATOR_MODE_BASE_RIGHT)
286 switch_buffer_by_offset (wordl0, wordl1, wordl2, wordl3, combs_buf[0].pw_len);
293 const u32 nr_len = ikepsk_bufs[salt_pos].nr_len;
294 const u32 msg_len = ikepsk_bufs[salt_pos].msg_len;
300 w_s[lid] = swap32 (ikepsk_bufs[salt_pos].nr_buf[lid]);
303 barrier (CLK_LOCAL_MEM_FENCE);
305 __local u32 s_msg_buf[128];
307 const u32 lid2 = lid * 2;
309 s_msg_buf[lid2 + 0] = swap32 (ikepsk_bufs[salt_pos].msg_buf[lid2 + 0]);
310 s_msg_buf[lid2 + 1] = swap32 (ikepsk_bufs[salt_pos].msg_buf[lid2 + 1]);
312 barrier (CLK_LOCAL_MEM_FENCE);
314 if (gid >= gid_max) return;
320 for (u32 il_pos = 0; il_pos < combs_cnt; il_pos++)
322 const u32 pw_r_len = combs_buf[il_pos].pw_len;
324 const u32 pw_len = pw_l_len + pw_r_len;
328 wordr0[0] = combs_buf[il_pos].i[0];
329 wordr0[1] = combs_buf[il_pos].i[1];
330 wordr0[2] = combs_buf[il_pos].i[2];
331 wordr0[3] = combs_buf[il_pos].i[3];
335 wordr1[0] = combs_buf[il_pos].i[4];
336 wordr1[1] = combs_buf[il_pos].i[5];
337 wordr1[2] = combs_buf[il_pos].i[6];
338 wordr1[3] = combs_buf[il_pos].i[7];
354 if (combs_mode == COMBINATOR_MODE_BASE_LEFT)
356 switch_buffer_by_offset (wordr0, wordr1, wordr2, wordr3, pw_l_len);
361 w0[0] = wordl0[0] | wordr0[0];
362 w0[1] = wordl0[1] | wordr0[1];
363 w0[2] = wordl0[2] | wordr0[2];
364 w0[3] = wordl0[3] | wordr0[3];
368 w1[0] = wordl1[0] | wordr1[0];
369 w1[1] = wordl1[1] | wordr1[1];
370 w1[2] = wordl1[2] | wordr1[2];
371 w1[3] = wordl1[3] | wordr1[3];
375 w2[0] = wordl2[0] | wordr2[0];
376 w2[1] = wordl2[1] | wordr2[1];
377 w2[2] = wordl2[2] | wordr2[2];
378 w2[3] = wordl2[3] | wordr2[3];
382 w3[0] = wordl3[0] | wordr3[0];
383 w3[1] = wordl3[1] | wordr3[1];
384 w3[2] = wordl3[2] | wordr3[2];
385 w3[3] = wordl3[3] | wordr3[3];
393 w0_t[0] = swap32 (w0[0]);
394 w0_t[1] = swap32 (w0[1]);
395 w0_t[2] = swap32 (w0[2]);
396 w0_t[3] = swap32 (w0[3]);
400 w1_t[0] = swap32 (w1[0]);
401 w1_t[1] = swap32 (w1[1]);
402 w1_t[2] = swap32 (w1[2]);
403 w1_t[3] = swap32 (w1[3]);
422 hmac_sha1_pad (w0_t, w1_t, w2_t, w3_t, ipad, opad);
439 w3_t[3] = (64 + nr_len) * 8;
443 hmac_sha1_run (w0_t, w1_t, w2_t, w3_t, ipad, opad, digest);
462 hmac_sha1_pad (w0_t, w1_t, w2_t, w3_t, ipad, opad);
467 for (left = ikepsk_bufs[salt_pos].msg_len, off = 0; left >= 56; left -= 64, off += 16)
469 w0_t[0] = s_msg_buf[off + 0];
470 w0_t[1] = s_msg_buf[off + 1];
471 w0_t[2] = s_msg_buf[off + 2];
472 w0_t[3] = s_msg_buf[off + 3];
473 w1_t[0] = s_msg_buf[off + 4];
474 w1_t[1] = s_msg_buf[off + 5];
475 w1_t[2] = s_msg_buf[off + 6];
476 w1_t[3] = s_msg_buf[off + 7];
477 w2_t[0] = s_msg_buf[off + 8];
478 w2_t[1] = s_msg_buf[off + 9];
479 w2_t[2] = s_msg_buf[off + 10];
480 w2_t[3] = s_msg_buf[off + 11];
481 w3_t[0] = s_msg_buf[off + 12];
482 w3_t[1] = s_msg_buf[off + 13];
483 w3_t[2] = s_msg_buf[off + 14];
484 w3_t[3] = s_msg_buf[off + 15];
486 sha1_transform (w0_t, w1_t, w2_t, w3_t, ipad);
489 w0_t[0] = s_msg_buf[off + 0];
490 w0_t[1] = s_msg_buf[off + 1];
491 w0_t[2] = s_msg_buf[off + 2];
492 w0_t[3] = s_msg_buf[off + 3];
493 w1_t[0] = s_msg_buf[off + 4];
494 w1_t[1] = s_msg_buf[off + 5];
495 w1_t[2] = s_msg_buf[off + 6];
496 w1_t[3] = s_msg_buf[off + 7];
497 w2_t[0] = s_msg_buf[off + 8];
498 w2_t[1] = s_msg_buf[off + 9];
499 w2_t[2] = s_msg_buf[off + 10];
500 w2_t[3] = s_msg_buf[off + 11];
501 w3_t[0] = s_msg_buf[off + 12];
502 w3_t[1] = s_msg_buf[off + 13];
504 w3_t[3] = (64 + msg_len) * 8;
506 hmac_sha1_run (w0_t, w1_t, w2_t, w3_t, ipad, opad, digest);
508 const u32 r0 = digest[3];
509 const u32 r1 = digest[4];
510 const u32 r2 = digest[2];
511 const u32 r3 = digest[1];
517 __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)
521 __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)
525 __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 combs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
531 const u32 lid = get_local_id (0);
537 const u32 gid = get_global_id (0);
541 wordl0[0] = pws[gid].i[ 0];
542 wordl0[1] = pws[gid].i[ 1];
543 wordl0[2] = pws[gid].i[ 2];
544 wordl0[3] = pws[gid].i[ 3];
548 wordl1[0] = pws[gid].i[ 4];
549 wordl1[1] = pws[gid].i[ 5];
550 wordl1[2] = pws[gid].i[ 6];
551 wordl1[3] = pws[gid].i[ 7];
567 const u32 pw_l_len = pws[gid].pw_len;
569 if (combs_mode == COMBINATOR_MODE_BASE_RIGHT)
571 switch_buffer_by_offset (wordl0, wordl1, wordl2, wordl3, combs_buf[0].pw_len);
578 const u32 nr_len = ikepsk_bufs[salt_pos].nr_len;
579 const u32 msg_len = ikepsk_bufs[salt_pos].msg_len;
585 w_s[lid] = swap32 (ikepsk_bufs[salt_pos].nr_buf[lid]);
588 barrier (CLK_LOCAL_MEM_FENCE);
590 __local u32 s_msg_buf[128];
592 const u32 lid2 = lid * 2;
594 s_msg_buf[lid2 + 0] = swap32 (ikepsk_bufs[salt_pos].msg_buf[lid2 + 0]);
595 s_msg_buf[lid2 + 1] = swap32 (ikepsk_bufs[salt_pos].msg_buf[lid2 + 1]);
597 barrier (CLK_LOCAL_MEM_FENCE);
599 if (gid >= gid_max) return;
605 const u32 search[4] =
607 digests_buf[digests_offset].digest_buf[DGST_R0],
608 digests_buf[digests_offset].digest_buf[DGST_R1],
609 digests_buf[digests_offset].digest_buf[DGST_R2],
610 digests_buf[digests_offset].digest_buf[DGST_R3]
617 for (u32 il_pos = 0; il_pos < combs_cnt; il_pos++)
619 const u32 pw_r_len = combs_buf[il_pos].pw_len;
621 const u32 pw_len = pw_l_len + pw_r_len;
625 wordr0[0] = combs_buf[il_pos].i[0];
626 wordr0[1] = combs_buf[il_pos].i[1];
627 wordr0[2] = combs_buf[il_pos].i[2];
628 wordr0[3] = combs_buf[il_pos].i[3];
632 wordr1[0] = combs_buf[il_pos].i[4];
633 wordr1[1] = combs_buf[il_pos].i[5];
634 wordr1[2] = combs_buf[il_pos].i[6];
635 wordr1[3] = combs_buf[il_pos].i[7];
651 if (combs_mode == COMBINATOR_MODE_BASE_LEFT)
653 switch_buffer_by_offset (wordr0, wordr1, wordr2, wordr3, pw_l_len);
658 w0[0] = wordl0[0] | wordr0[0];
659 w0[1] = wordl0[1] | wordr0[1];
660 w0[2] = wordl0[2] | wordr0[2];
661 w0[3] = wordl0[3] | wordr0[3];
665 w1[0] = wordl1[0] | wordr1[0];
666 w1[1] = wordl1[1] | wordr1[1];
667 w1[2] = wordl1[2] | wordr1[2];
668 w1[3] = wordl1[3] | wordr1[3];
672 w2[0] = wordl2[0] | wordr2[0];
673 w2[1] = wordl2[1] | wordr2[1];
674 w2[2] = wordl2[2] | wordr2[2];
675 w2[3] = wordl2[3] | wordr2[3];
679 w3[0] = wordl3[0] | wordr3[0];
680 w3[1] = wordl3[1] | wordr3[1];
681 w3[2] = wordl3[2] | wordr3[2];
682 w3[3] = wordl3[3] | wordr3[3];
690 w0_t[0] = swap32 (w0[0]);
691 w0_t[1] = swap32 (w0[1]);
692 w0_t[2] = swap32 (w0[2]);
693 w0_t[3] = swap32 (w0[3]);
697 w1_t[0] = swap32 (w1[0]);
698 w1_t[1] = swap32 (w1[1]);
699 w1_t[2] = swap32 (w1[2]);
700 w1_t[3] = swap32 (w1[3]);
719 hmac_sha1_pad (w0_t, w1_t, w2_t, w3_t, ipad, opad);
736 w3_t[3] = (64 + nr_len) * 8;
740 hmac_sha1_run (w0_t, w1_t, w2_t, w3_t, ipad, opad, digest);
759 hmac_sha1_pad (w0_t, w1_t, w2_t, w3_t, ipad, opad);
764 for (left = ikepsk_bufs[salt_pos].msg_len, off = 0; left >= 56; left -= 64, off += 16)
766 w0_t[0] = s_msg_buf[off + 0];
767 w0_t[1] = s_msg_buf[off + 1];
768 w0_t[2] = s_msg_buf[off + 2];
769 w0_t[3] = s_msg_buf[off + 3];
770 w1_t[0] = s_msg_buf[off + 4];
771 w1_t[1] = s_msg_buf[off + 5];
772 w1_t[2] = s_msg_buf[off + 6];
773 w1_t[3] = s_msg_buf[off + 7];
774 w2_t[0] = s_msg_buf[off + 8];
775 w2_t[1] = s_msg_buf[off + 9];
776 w2_t[2] = s_msg_buf[off + 10];
777 w2_t[3] = s_msg_buf[off + 11];
778 w3_t[0] = s_msg_buf[off + 12];
779 w3_t[1] = s_msg_buf[off + 13];
780 w3_t[2] = s_msg_buf[off + 14];
781 w3_t[3] = s_msg_buf[off + 15];
783 sha1_transform (w0_t, w1_t, w2_t, w3_t, ipad);
786 w0_t[0] = s_msg_buf[off + 0];
787 w0_t[1] = s_msg_buf[off + 1];
788 w0_t[2] = s_msg_buf[off + 2];
789 w0_t[3] = s_msg_buf[off + 3];
790 w1_t[0] = s_msg_buf[off + 4];
791 w1_t[1] = s_msg_buf[off + 5];
792 w1_t[2] = s_msg_buf[off + 6];
793 w1_t[3] = s_msg_buf[off + 7];
794 w2_t[0] = s_msg_buf[off + 8];
795 w2_t[1] = s_msg_buf[off + 9];
796 w2_t[2] = s_msg_buf[off + 10];
797 w2_t[3] = s_msg_buf[off + 11];
798 w3_t[0] = s_msg_buf[off + 12];
799 w3_t[1] = s_msg_buf[off + 13];
801 w3_t[3] = (64 + msg_len) * 8;
803 hmac_sha1_run (w0_t, w1_t, w2_t, w3_t, ipad, opad, digest);
805 const u32 r0 = digest[3];
806 const u32 r1 = digest[4];
807 const u32 r2 = digest[2];
808 const u32 r3 = digest[1];
814 __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)
818 __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)