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 "types_ocl.c"
21 #define COMPARE_M "check_multi_vect1_comp4.c"
25 #define COMPARE_M "check_multi_vect2_comp4.c"
29 #define COMPARE_M "check_multi_vect4_comp4.c"
32 __constant u32 k_sha256[64] =
34 SHA256C00, SHA256C01, SHA256C02, SHA256C03,
35 SHA256C04, SHA256C05, SHA256C06, SHA256C07,
36 SHA256C08, SHA256C09, SHA256C0a, SHA256C0b,
37 SHA256C0c, SHA256C0d, SHA256C0e, SHA256C0f,
38 SHA256C10, SHA256C11, SHA256C12, SHA256C13,
39 SHA256C14, SHA256C15, SHA256C16, SHA256C17,
40 SHA256C18, SHA256C19, SHA256C1a, SHA256C1b,
41 SHA256C1c, SHA256C1d, SHA256C1e, SHA256C1f,
42 SHA256C20, SHA256C21, SHA256C22, SHA256C23,
43 SHA256C24, SHA256C25, SHA256C26, SHA256C27,
44 SHA256C28, SHA256C29, SHA256C2a, SHA256C2b,
45 SHA256C2c, SHA256C2d, SHA256C2e, SHA256C2f,
46 SHA256C30, SHA256C31, SHA256C32, SHA256C33,
47 SHA256C34, SHA256C35, SHA256C36, SHA256C37,
48 SHA256C38, SHA256C39, SHA256C3a, SHA256C3b,
49 SHA256C3c, SHA256C3d, SHA256C3e, SHA256C3f,
52 static void sha256_transform (const u32 w0[4], const u32 w1[4], const u32 w2[4], const u32 w3[4], u32 digest[8])
80 #define ROUND_EXPAND() \
82 w0_t = SHA256_EXPAND (we_t, w9_t, w1_t, w0_t); \
83 w1_t = SHA256_EXPAND (wf_t, wa_t, w2_t, w1_t); \
84 w2_t = SHA256_EXPAND (w0_t, wb_t, w3_t, w2_t); \
85 w3_t = SHA256_EXPAND (w1_t, wc_t, w4_t, w3_t); \
86 w4_t = SHA256_EXPAND (w2_t, wd_t, w5_t, w4_t); \
87 w5_t = SHA256_EXPAND (w3_t, we_t, w6_t, w5_t); \
88 w6_t = SHA256_EXPAND (w4_t, wf_t, w7_t, w6_t); \
89 w7_t = SHA256_EXPAND (w5_t, w0_t, w8_t, w7_t); \
90 w8_t = SHA256_EXPAND (w6_t, w1_t, w9_t, w8_t); \
91 w9_t = SHA256_EXPAND (w7_t, w2_t, wa_t, w9_t); \
92 wa_t = SHA256_EXPAND (w8_t, w3_t, wb_t, wa_t); \
93 wb_t = SHA256_EXPAND (w9_t, w4_t, wc_t, wb_t); \
94 wc_t = SHA256_EXPAND (wa_t, w5_t, wd_t, wc_t); \
95 wd_t = SHA256_EXPAND (wb_t, w6_t, we_t, wd_t); \
96 we_t = SHA256_EXPAND (wc_t, w7_t, wf_t, we_t); \
97 wf_t = SHA256_EXPAND (wd_t, w8_t, w0_t, wf_t); \
100 #define ROUND_STEP(i) \
102 SHA256_STEP (SHA256_F0o, SHA256_F1o, a, b, c, d, e, f, g, h, w0_t, k_sha256[i + 0]); \
103 SHA256_STEP (SHA256_F0o, SHA256_F1o, h, a, b, c, d, e, f, g, w1_t, k_sha256[i + 1]); \
104 SHA256_STEP (SHA256_F0o, SHA256_F1o, g, h, a, b, c, d, e, f, w2_t, k_sha256[i + 2]); \
105 SHA256_STEP (SHA256_F0o, SHA256_F1o, f, g, h, a, b, c, d, e, w3_t, k_sha256[i + 3]); \
106 SHA256_STEP (SHA256_F0o, SHA256_F1o, e, f, g, h, a, b, c, d, w4_t, k_sha256[i + 4]); \
107 SHA256_STEP (SHA256_F0o, SHA256_F1o, d, e, f, g, h, a, b, c, w5_t, k_sha256[i + 5]); \
108 SHA256_STEP (SHA256_F0o, SHA256_F1o, c, d, e, f, g, h, a, b, w6_t, k_sha256[i + 6]); \
109 SHA256_STEP (SHA256_F0o, SHA256_F1o, b, c, d, e, f, g, h, a, w7_t, k_sha256[i + 7]); \
110 SHA256_STEP (SHA256_F0o, SHA256_F1o, a, b, c, d, e, f, g, h, w8_t, k_sha256[i + 8]); \
111 SHA256_STEP (SHA256_F0o, SHA256_F1o, h, a, b, c, d, e, f, g, w9_t, k_sha256[i + 9]); \
112 SHA256_STEP (SHA256_F0o, SHA256_F1o, g, h, a, b, c, d, e, f, wa_t, k_sha256[i + 10]); \
113 SHA256_STEP (SHA256_F0o, SHA256_F1o, f, g, h, a, b, c, d, e, wb_t, k_sha256[i + 11]); \
114 SHA256_STEP (SHA256_F0o, SHA256_F1o, e, f, g, h, a, b, c, d, wc_t, k_sha256[i + 12]); \
115 SHA256_STEP (SHA256_F0o, SHA256_F1o, d, e, f, g, h, a, b, c, wd_t, k_sha256[i + 13]); \
116 SHA256_STEP (SHA256_F0o, SHA256_F1o, c, d, e, f, g, h, a, b, we_t, k_sha256[i + 14]); \
117 SHA256_STEP (SHA256_F0o, SHA256_F1o, b, c, d, e, f, g, h, a, wf_t, k_sha256[i + 15]); \
123 for (int i = 16; i < 64; i += 16)
125 ROUND_EXPAND (); ROUND_STEP (i);
138 __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m05200_init (__global pw_t *pws, __global gpu_rule_t *rules_buf, __global comb_t *combs_buf, __global bf_t *bfs_buf, __global pwsafe3_tmp_t *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 void *esalt_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)
144 const u32 gid = get_global_id (0);
146 if (gid >= gid_max) return;
150 w0[0] = pws[gid].i[ 0];
151 w0[1] = pws[gid].i[ 1];
152 w0[2] = pws[gid].i[ 2];
153 w0[3] = pws[gid].i[ 3];
157 w1[0] = pws[gid].i[ 4];
158 w1[1] = pws[gid].i[ 5];
176 const u32 pw_len = pws[gid].pw_len;
184 salt_buf0[0] = salt_bufs[salt_pos].salt_buf[0];
185 salt_buf0[1] = salt_bufs[salt_pos].salt_buf[1];
186 salt_buf0[2] = salt_bufs[salt_pos].salt_buf[2];
187 salt_buf0[3] = salt_bufs[salt_pos].salt_buf[3];
191 salt_buf1[0] = salt_bufs[salt_pos].salt_buf[4];
192 salt_buf1[1] = salt_bufs[salt_pos].salt_buf[5];
193 salt_buf1[2] = salt_bufs[salt_pos].salt_buf[6];
194 salt_buf1[3] = salt_bufs[salt_pos].salt_buf[7];
210 u32 salt_len = salt_bufs[salt_pos].salt_len;
212 switch_buffer_by_offset (salt_buf0, salt_buf1, salt_buf2, salt_buf3, pw_len);
214 w0[0] |= salt_buf0[0];
215 w0[1] |= salt_buf0[1];
216 w0[2] |= salt_buf0[2];
217 w0[3] |= salt_buf0[3];
219 w1[0] |= salt_buf1[0];
220 w1[1] |= salt_buf1[1];
221 w1[2] |= salt_buf1[2];
222 w1[3] |= salt_buf1[3];
224 w2[0] |= salt_buf2[0];
225 w2[1] |= salt_buf2[1];
226 w2[2] |= salt_buf2[2];
227 w2[3] |= salt_buf2[3];
229 w3[0] |= salt_buf3[0];
230 w3[1] |= salt_buf3[1];
231 w3[2] |= salt_buf3[2];
232 w3[3] |= salt_buf3[3];
234 const u32 block_len = pw_len + salt_len;
236 append_0x80_4x4 (w0, w1, w2, w3, block_len);
242 w0[0] = swap_workaround (w0[0]);
243 w0[1] = swap_workaround (w0[1]);
244 w0[2] = swap_workaround (w0[2]);
245 w0[3] = swap_workaround (w0[3]);
247 w1[0] = swap_workaround (w1[0]);
248 w1[1] = swap_workaround (w1[1]);
249 w1[2] = swap_workaround (w1[2]);
250 w1[3] = swap_workaround (w1[3]);
252 w2[0] = swap_workaround (w2[0]);
253 w2[1] = swap_workaround (w2[1]);
254 w2[2] = swap_workaround (w2[2]);
255 w2[3] = swap_workaround (w2[3]);
257 w3[0] = swap_workaround (w3[0]);
258 w3[1] = swap_workaround (w3[1]);
260 w3[3] = block_len * 8;
268 digest[0] = SHA256M_A;
269 digest[1] = SHA256M_B;
270 digest[2] = SHA256M_C;
271 digest[3] = SHA256M_D;
272 digest[4] = SHA256M_E;
273 digest[5] = SHA256M_F;
274 digest[6] = SHA256M_G;
275 digest[7] = SHA256M_H;
277 sha256_transform (w0, w1, w2, w3, digest);
283 tmps[gid].digest_buf[0] = digest[0];
284 tmps[gid].digest_buf[1] = digest[1];
285 tmps[gid].digest_buf[2] = digest[2];
286 tmps[gid].digest_buf[3] = digest[3];
287 tmps[gid].digest_buf[4] = digest[4];
288 tmps[gid].digest_buf[5] = digest[5];
289 tmps[gid].digest_buf[6] = digest[6];
290 tmps[gid].digest_buf[7] = digest[7];
293 __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m05200_loop (__global pw_t *pws, __global gpu_rule_t *rules_buf, __global comb_t *combs_buf, __global bf_t *bfs_buf, __global pwsafe3_tmp_t *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 void *esalt_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)
295 const u32 gid = get_global_id (0);
297 if (gid >= gid_max) return;
301 digest[0] = tmps[gid].digest_buf[0];
302 digest[1] = tmps[gid].digest_buf[1];
303 digest[2] = tmps[gid].digest_buf[2];
304 digest[3] = tmps[gid].digest_buf[3];
305 digest[4] = tmps[gid].digest_buf[4];
306 digest[5] = tmps[gid].digest_buf[5];
307 digest[6] = tmps[gid].digest_buf[6];
308 digest[7] = tmps[gid].digest_buf[7];
314 for (u32 i = 0; i < loop_cnt; i++)
344 digest[0] = SHA256M_A;
345 digest[1] = SHA256M_B;
346 digest[2] = SHA256M_C;
347 digest[3] = SHA256M_D;
348 digest[4] = SHA256M_E;
349 digest[5] = SHA256M_F;
350 digest[6] = SHA256M_G;
351 digest[7] = SHA256M_H;
353 sha256_transform (w0, w1, w2, w3, digest);
356 tmps[gid].digest_buf[0] = digest[0];
357 tmps[gid].digest_buf[1] = digest[1];
358 tmps[gid].digest_buf[2] = digest[2];
359 tmps[gid].digest_buf[3] = digest[3];
360 tmps[gid].digest_buf[4] = digest[4];
361 tmps[gid].digest_buf[5] = digest[5];
362 tmps[gid].digest_buf[6] = digest[6];
363 tmps[gid].digest_buf[7] = digest[7];
366 __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m05200_comp (__global pw_t *pws, __global gpu_rule_t *rules_buf, __global comb_t *combs_buf, __global bf_t *bfs_buf, __global pwsafe3_tmp_t *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 void *esalt_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)
372 const u32 gid = get_global_id (0);
374 if (gid >= gid_max) return;
376 const u32 lid = get_local_id (0);
382 const u32 r0 = tmps[gid].digest_buf[DGST_R0];
383 const u32 r1 = tmps[gid].digest_buf[DGST_R1];
384 const u32 r2 = tmps[gid].digest_buf[DGST_R2];
385 const u32 r3 = tmps[gid].digest_buf[DGST_R3];