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"
19 #include "include/rp_gpu.h"
22 #define COMPARE_S "OpenCL/check_single_comp4.c"
23 #define COMPARE_M "OpenCL/check_multi_comp4.c"
25 static void sha1_transform (const u32 w0[4], const u32 w1[4], const u32 w2[4], const u32 w3[4], u32 digest[5])
53 SHA1_STEP (SHA1_F0o, A, B, C, D, E, w0_t);
54 SHA1_STEP (SHA1_F0o, E, A, B, C, D, w1_t);
55 SHA1_STEP (SHA1_F0o, D, E, A, B, C, w2_t);
56 SHA1_STEP (SHA1_F0o, C, D, E, A, B, w3_t);
57 SHA1_STEP (SHA1_F0o, B, C, D, E, A, w4_t);
58 SHA1_STEP (SHA1_F0o, A, B, C, D, E, w5_t);
59 SHA1_STEP (SHA1_F0o, E, A, B, C, D, w6_t);
60 SHA1_STEP (SHA1_F0o, D, E, A, B, C, w7_t);
61 SHA1_STEP (SHA1_F0o, C, D, E, A, B, w8_t);
62 SHA1_STEP (SHA1_F0o, B, C, D, E, A, w9_t);
63 SHA1_STEP (SHA1_F0o, A, B, C, D, E, wa_t);
64 SHA1_STEP (SHA1_F0o, E, A, B, C, D, wb_t);
65 SHA1_STEP (SHA1_F0o, D, E, A, B, C, wc_t);
66 SHA1_STEP (SHA1_F0o, C, D, E, A, B, wd_t);
67 SHA1_STEP (SHA1_F0o, B, C, D, E, A, we_t);
68 SHA1_STEP (SHA1_F0o, A, B, C, D, E, wf_t);
69 w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F0o, E, A, B, C, D, w0_t);
70 w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F0o, D, E, A, B, C, w1_t);
71 w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F0o, C, D, E, A, B, w2_t);
72 w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F0o, B, C, D, E, A, w3_t);
77 w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w4_t);
78 w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w5_t);
79 w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w6_t);
80 w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w7_t);
81 w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w8_t);
82 w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w9_t);
83 wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wa_t);
84 wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, wb_t);
85 wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, wc_t);
86 wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, wd_t);
87 we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, we_t);
88 wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wf_t);
89 w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w0_t);
90 w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w1_t);
91 w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w2_t);
92 w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w3_t);
93 w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w4_t);
94 w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w5_t);
95 w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w6_t);
96 w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w7_t);
101 w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, w8_t);
102 w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, w9_t);
103 wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, wa_t);
104 wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, wb_t);
105 wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, wc_t);
106 wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, wd_t);
107 we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, we_t);
108 wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, wf_t);
109 w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, w0_t);
110 w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, w1_t);
111 w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, w2_t);
112 w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, w3_t);
113 w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, w4_t);
114 w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, w5_t);
115 w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, w6_t);
116 w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, w7_t);
117 w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, w8_t);
118 w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, w9_t);
119 wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, wa_t);
120 wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, wb_t);
125 wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, wc_t);
126 wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wd_t);
127 we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, we_t);
128 wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, wf_t);
129 w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w0_t);
130 w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w1_t);
131 w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w2_t);
132 w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w3_t);
133 w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w4_t);
134 w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w5_t);
135 w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w6_t);
136 w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w7_t);
137 w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w8_t);
138 w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w9_t);
139 wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, wa_t);
140 wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, wb_t);
141 wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wc_t);
142 wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, wd_t);
143 we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, we_t);
144 wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, wf_t);
153 static void hmac_sha1_pad (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], u32 ipad[5], u32 opad[5])
155 w0[0] = w0[0] ^ 0x36363636;
156 w0[1] = w0[1] ^ 0x36363636;
157 w0[2] = w0[2] ^ 0x36363636;
158 w0[3] = w0[3] ^ 0x36363636;
159 w1[0] = w1[0] ^ 0x36363636;
160 w1[1] = w1[1] ^ 0x36363636;
161 w1[2] = w1[2] ^ 0x36363636;
162 w1[3] = w1[3] ^ 0x36363636;
163 w2[0] = w2[0] ^ 0x36363636;
164 w2[1] = w2[1] ^ 0x36363636;
165 w2[2] = w2[2] ^ 0x36363636;
166 w2[3] = w2[3] ^ 0x36363636;
167 w3[0] = w3[0] ^ 0x36363636;
168 w3[1] = w3[1] ^ 0x36363636;
169 w3[2] = w3[2] ^ 0x36363636;
170 w3[3] = w3[3] ^ 0x36363636;
178 sha1_transform (w0, w1, w2, w3, ipad);
180 w0[0] = w0[0] ^ 0x6a6a6a6a;
181 w0[1] = w0[1] ^ 0x6a6a6a6a;
182 w0[2] = w0[2] ^ 0x6a6a6a6a;
183 w0[3] = w0[3] ^ 0x6a6a6a6a;
184 w1[0] = w1[0] ^ 0x6a6a6a6a;
185 w1[1] = w1[1] ^ 0x6a6a6a6a;
186 w1[2] = w1[2] ^ 0x6a6a6a6a;
187 w1[3] = w1[3] ^ 0x6a6a6a6a;
188 w2[0] = w2[0] ^ 0x6a6a6a6a;
189 w2[1] = w2[1] ^ 0x6a6a6a6a;
190 w2[2] = w2[2] ^ 0x6a6a6a6a;
191 w2[3] = w2[3] ^ 0x6a6a6a6a;
192 w3[0] = w3[0] ^ 0x6a6a6a6a;
193 w3[1] = w3[1] ^ 0x6a6a6a6a;
194 w3[2] = w3[2] ^ 0x6a6a6a6a;
195 w3[3] = w3[3] ^ 0x6a6a6a6a;
203 sha1_transform (w0, w1, w2, w3, opad);
206 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])
214 sha1_transform (w0, w1, w2, w3, digest);
231 w3[3] = (64 + 20) * 8;
239 sha1_transform (w0, w1, w2, w3, digest);
242 __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)
248 const u32 lid = get_local_id (0);
254 const u32 gid = get_global_id (0);
258 pw_buf0[0] = pws[gid].i[ 0];
259 pw_buf0[1] = pws[gid].i[ 1];
260 pw_buf0[2] = pws[gid].i[ 2];
261 pw_buf0[3] = pws[gid].i[ 3];
265 pw_buf1[0] = pws[gid].i[ 4];
266 pw_buf1[1] = pws[gid].i[ 5];
267 pw_buf1[2] = pws[gid].i[ 6];
268 pw_buf1[3] = pws[gid].i[ 7];
270 const u32 pw_len = pws[gid].pw_len;
276 const u32 nr_len = ikepsk_bufs[salt_pos].nr_len;
277 const u32 msg_len = ikepsk_bufs[salt_pos].msg_len;
283 w_s[lid] = swap32 (ikepsk_bufs[salt_pos].nr_buf[lid]);
286 barrier (CLK_LOCAL_MEM_FENCE);
288 __local u32 s_msg_buf[128];
290 const u32 lid2 = lid * 2;
292 s_msg_buf[lid2 + 0] = swap32 (ikepsk_bufs[salt_pos].msg_buf[lid2 + 0]);
293 s_msg_buf[lid2 + 1] = swap32 (ikepsk_bufs[salt_pos].msg_buf[lid2 + 1]);
295 barrier (CLK_LOCAL_MEM_FENCE);
297 if (gid >= gid_max) return;
303 for (u32 il_pos = 0; il_pos < rules_cnt; il_pos++)
333 const u32 out_len = apply_rules (rules_buf[il_pos].cmds, w0, w1, pw_len);
341 w0_t[0] = swap32 (w0[0]);
342 w0_t[1] = swap32 (w0[1]);
343 w0_t[2] = swap32 (w0[2]);
344 w0_t[3] = swap32 (w0[3]);
348 w1_t[0] = swap32 (w1[0]);
349 w1_t[1] = swap32 (w1[1]);
350 w1_t[2] = swap32 (w1[2]);
351 w1_t[3] = swap32 (w1[3]);
370 hmac_sha1_pad (w0_t, w1_t, w2_t, w3_t, ipad, opad);
387 w3_t[3] = (64 + nr_len) * 8;
391 hmac_sha1_run (w0_t, w1_t, w2_t, w3_t, ipad, opad, digest);
410 hmac_sha1_pad (w0_t, w1_t, w2_t, w3_t, ipad, opad);
415 for (left = ikepsk_bufs[salt_pos].msg_len, off = 0; left >= 56; left -= 64, off += 16)
417 w0_t[0] = s_msg_buf[off + 0];
418 w0_t[1] = s_msg_buf[off + 1];
419 w0_t[2] = s_msg_buf[off + 2];
420 w0_t[3] = s_msg_buf[off + 3];
421 w1_t[0] = s_msg_buf[off + 4];
422 w1_t[1] = s_msg_buf[off + 5];
423 w1_t[2] = s_msg_buf[off + 6];
424 w1_t[3] = s_msg_buf[off + 7];
425 w2_t[0] = s_msg_buf[off + 8];
426 w2_t[1] = s_msg_buf[off + 9];
427 w2_t[2] = s_msg_buf[off + 10];
428 w2_t[3] = s_msg_buf[off + 11];
429 w3_t[0] = s_msg_buf[off + 12];
430 w3_t[1] = s_msg_buf[off + 13];
431 w3_t[2] = s_msg_buf[off + 14];
432 w3_t[3] = s_msg_buf[off + 15];
434 sha1_transform (w0_t, w1_t, w2_t, w3_t, ipad);
437 w0_t[0] = s_msg_buf[off + 0];
438 w0_t[1] = s_msg_buf[off + 1];
439 w0_t[2] = s_msg_buf[off + 2];
440 w0_t[3] = s_msg_buf[off + 3];
441 w1_t[0] = s_msg_buf[off + 4];
442 w1_t[1] = s_msg_buf[off + 5];
443 w1_t[2] = s_msg_buf[off + 6];
444 w1_t[3] = s_msg_buf[off + 7];
445 w2_t[0] = s_msg_buf[off + 8];
446 w2_t[1] = s_msg_buf[off + 9];
447 w2_t[2] = s_msg_buf[off + 10];
448 w2_t[3] = s_msg_buf[off + 11];
449 w3_t[0] = s_msg_buf[off + 12];
450 w3_t[1] = s_msg_buf[off + 13];
452 w3_t[3] = (64 + msg_len) * 8;
454 hmac_sha1_run (w0_t, w1_t, w2_t, w3_t, ipad, opad, digest);
456 const u32 r0 = digest[3];
457 const u32 r1 = digest[4];
458 const u32 r2 = digest[2];
459 const u32 r3 = digest[1];
465 __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)
469 __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)
473 __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)
479 const u32 lid = get_local_id (0);
485 const u32 gid = get_global_id (0);
489 pw_buf0[0] = pws[gid].i[ 0];
490 pw_buf0[1] = pws[gid].i[ 1];
491 pw_buf0[2] = pws[gid].i[ 2];
492 pw_buf0[3] = pws[gid].i[ 3];
496 pw_buf1[0] = pws[gid].i[ 4];
497 pw_buf1[1] = pws[gid].i[ 5];
498 pw_buf1[2] = pws[gid].i[ 6];
499 pw_buf1[3] = pws[gid].i[ 7];
501 const u32 pw_len = pws[gid].pw_len;
507 const u32 nr_len = ikepsk_bufs[salt_pos].nr_len;
508 const u32 msg_len = ikepsk_bufs[salt_pos].msg_len;
514 w_s[lid] = swap32 (ikepsk_bufs[salt_pos].nr_buf[lid]);
517 barrier (CLK_LOCAL_MEM_FENCE);
519 __local u32 s_msg_buf[128];
521 const u32 lid2 = lid * 2;
523 s_msg_buf[lid2 + 0] = swap32 (ikepsk_bufs[salt_pos].msg_buf[lid2 + 0]);
524 s_msg_buf[lid2 + 1] = swap32 (ikepsk_bufs[salt_pos].msg_buf[lid2 + 1]);
526 barrier (CLK_LOCAL_MEM_FENCE);
528 if (gid >= gid_max) return;
534 const u32 search[4] =
536 digests_buf[digests_offset].digest_buf[DGST_R0],
537 digests_buf[digests_offset].digest_buf[DGST_R1],
538 digests_buf[digests_offset].digest_buf[DGST_R2],
539 digests_buf[digests_offset].digest_buf[DGST_R3]
546 for (u32 il_pos = 0; il_pos < rules_cnt; il_pos++)
576 const u32 out_len = apply_rules (rules_buf[il_pos].cmds, w0, w1, pw_len);
584 w0_t[0] = swap32 (w0[0]);
585 w0_t[1] = swap32 (w0[1]);
586 w0_t[2] = swap32 (w0[2]);
587 w0_t[3] = swap32 (w0[3]);
591 w1_t[0] = swap32 (w1[0]);
592 w1_t[1] = swap32 (w1[1]);
593 w1_t[2] = swap32 (w1[2]);
594 w1_t[3] = swap32 (w1[3]);
613 hmac_sha1_pad (w0_t, w1_t, w2_t, w3_t, ipad, opad);
630 w3_t[3] = (64 + nr_len) * 8;
634 hmac_sha1_run (w0_t, w1_t, w2_t, w3_t, ipad, opad, digest);
653 hmac_sha1_pad (w0_t, w1_t, w2_t, w3_t, ipad, opad);
658 for (left = ikepsk_bufs[salt_pos].msg_len, off = 0; left >= 56; left -= 64, off += 16)
660 w0_t[0] = s_msg_buf[off + 0];
661 w0_t[1] = s_msg_buf[off + 1];
662 w0_t[2] = s_msg_buf[off + 2];
663 w0_t[3] = s_msg_buf[off + 3];
664 w1_t[0] = s_msg_buf[off + 4];
665 w1_t[1] = s_msg_buf[off + 5];
666 w1_t[2] = s_msg_buf[off + 6];
667 w1_t[3] = s_msg_buf[off + 7];
668 w2_t[0] = s_msg_buf[off + 8];
669 w2_t[1] = s_msg_buf[off + 9];
670 w2_t[2] = s_msg_buf[off + 10];
671 w2_t[3] = s_msg_buf[off + 11];
672 w3_t[0] = s_msg_buf[off + 12];
673 w3_t[1] = s_msg_buf[off + 13];
674 w3_t[2] = s_msg_buf[off + 14];
675 w3_t[3] = s_msg_buf[off + 15];
677 sha1_transform (w0_t, w1_t, w2_t, w3_t, ipad);
680 w0_t[0] = s_msg_buf[off + 0];
681 w0_t[1] = s_msg_buf[off + 1];
682 w0_t[2] = s_msg_buf[off + 2];
683 w0_t[3] = s_msg_buf[off + 3];
684 w1_t[0] = s_msg_buf[off + 4];
685 w1_t[1] = s_msg_buf[off + 5];
686 w1_t[2] = s_msg_buf[off + 6];
687 w1_t[3] = s_msg_buf[off + 7];
688 w2_t[0] = s_msg_buf[off + 8];
689 w2_t[1] = s_msg_buf[off + 9];
690 w2_t[2] = s_msg_buf[off + 10];
691 w2_t[3] = s_msg_buf[off + 11];
692 w3_t[0] = s_msg_buf[off + 12];
693 w3_t[1] = s_msg_buf[off + 13];
695 w3_t[3] = (64 + msg_len) * 8;
697 hmac_sha1_run (w0_t, w1_t, w2_t, w3_t, ipad, opad, digest);
699 const u32 r0 = digest[3];
700 const u32 r1 = digest[4];
701 const u32 r2 = digest[2];
702 const u32 r3 = digest[1];
708 __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)
712 __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)