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"
20 #define COMPARE_S "check_single_comp4.c"
21 #define COMPARE_M "check_multi_comp4.c"
23 #define md5apr1_magic0 0x72706124
24 #define md5apr1_magic1 0x00002431
26 static void md5_transform (const u32 w0[4], const u32 w1[4], const u32 w2[4], const u32 w3[4], u32 digest[4])
52 MD5_STEP (MD5_Fo, a, b, c, d, w0_t, MD5C00, MD5S00);
53 MD5_STEP (MD5_Fo, d, a, b, c, w1_t, MD5C01, MD5S01);
54 MD5_STEP (MD5_Fo, c, d, a, b, w2_t, MD5C02, MD5S02);
55 MD5_STEP (MD5_Fo, b, c, d, a, w3_t, MD5C03, MD5S03);
56 MD5_STEP (MD5_Fo, a, b, c, d, w4_t, MD5C04, MD5S00);
57 MD5_STEP (MD5_Fo, d, a, b, c, w5_t, MD5C05, MD5S01);
58 MD5_STEP (MD5_Fo, c, d, a, b, w6_t, MD5C06, MD5S02);
59 MD5_STEP (MD5_Fo, b, c, d, a, w7_t, MD5C07, MD5S03);
60 MD5_STEP (MD5_Fo, a, b, c, d, w8_t, MD5C08, MD5S00);
61 MD5_STEP (MD5_Fo, d, a, b, c, w9_t, MD5C09, MD5S01);
62 MD5_STEP (MD5_Fo, c, d, a, b, wa_t, MD5C0a, MD5S02);
63 MD5_STEP (MD5_Fo, b, c, d, a, wb_t, MD5C0b, MD5S03);
64 MD5_STEP (MD5_Fo, a, b, c, d, wc_t, MD5C0c, MD5S00);
65 MD5_STEP (MD5_Fo, d, a, b, c, wd_t, MD5C0d, MD5S01);
66 MD5_STEP (MD5_Fo, c, d, a, b, we_t, MD5C0e, MD5S02);
67 MD5_STEP (MD5_Fo, b, c, d, a, wf_t, MD5C0f, MD5S03);
69 MD5_STEP (MD5_Go, a, b, c, d, w1_t, MD5C10, MD5S10);
70 MD5_STEP (MD5_Go, d, a, b, c, w6_t, MD5C11, MD5S11);
71 MD5_STEP (MD5_Go, c, d, a, b, wb_t, MD5C12, MD5S12);
72 MD5_STEP (MD5_Go, b, c, d, a, w0_t, MD5C13, MD5S13);
73 MD5_STEP (MD5_Go, a, b, c, d, w5_t, MD5C14, MD5S10);
74 MD5_STEP (MD5_Go, d, a, b, c, wa_t, MD5C15, MD5S11);
75 MD5_STEP (MD5_Go, c, d, a, b, wf_t, MD5C16, MD5S12);
76 MD5_STEP (MD5_Go, b, c, d, a, w4_t, MD5C17, MD5S13);
77 MD5_STEP (MD5_Go, a, b, c, d, w9_t, MD5C18, MD5S10);
78 MD5_STEP (MD5_Go, d, a, b, c, we_t, MD5C19, MD5S11);
79 MD5_STEP (MD5_Go, c, d, a, b, w3_t, MD5C1a, MD5S12);
80 MD5_STEP (MD5_Go, b, c, d, a, w8_t, MD5C1b, MD5S13);
81 MD5_STEP (MD5_Go, a, b, c, d, wd_t, MD5C1c, MD5S10);
82 MD5_STEP (MD5_Go, d, a, b, c, w2_t, MD5C1d, MD5S11);
83 MD5_STEP (MD5_Go, c, d, a, b, w7_t, MD5C1e, MD5S12);
84 MD5_STEP (MD5_Go, b, c, d, a, wc_t, MD5C1f, MD5S13);
86 MD5_STEP (MD5_H1, a, b, c, d, w5_t, MD5C20, MD5S20);
87 MD5_STEP (MD5_H2, d, a, b, c, w8_t, MD5C21, MD5S21);
88 MD5_STEP (MD5_H1, c, d, a, b, wb_t, MD5C22, MD5S22);
89 MD5_STEP (MD5_H2, b, c, d, a, we_t, MD5C23, MD5S23);
90 MD5_STEP (MD5_H1, a, b, c, d, w1_t, MD5C24, MD5S20);
91 MD5_STEP (MD5_H2, d, a, b, c, w4_t, MD5C25, MD5S21);
92 MD5_STEP (MD5_H1, c, d, a, b, w7_t, MD5C26, MD5S22);
93 MD5_STEP (MD5_H2, b, c, d, a, wa_t, MD5C27, MD5S23);
94 MD5_STEP (MD5_H1, a, b, c, d, wd_t, MD5C28, MD5S20);
95 MD5_STEP (MD5_H2, d, a, b, c, w0_t, MD5C29, MD5S21);
96 MD5_STEP (MD5_H1, c, d, a, b, w3_t, MD5C2a, MD5S22);
97 MD5_STEP (MD5_H2, b, c, d, a, w6_t, MD5C2b, MD5S23);
98 MD5_STEP (MD5_H1, a, b, c, d, w9_t, MD5C2c, MD5S20);
99 MD5_STEP (MD5_H2, d, a, b, c, wc_t, MD5C2d, MD5S21);
100 MD5_STEP (MD5_H1, c, d, a, b, wf_t, MD5C2e, MD5S22);
101 MD5_STEP (MD5_H2, b, c, d, a, w2_t, MD5C2f, MD5S23);
103 MD5_STEP (MD5_I , a, b, c, d, w0_t, MD5C30, MD5S30);
104 MD5_STEP (MD5_I , d, a, b, c, w7_t, MD5C31, MD5S31);
105 MD5_STEP (MD5_I , c, d, a, b, we_t, MD5C32, MD5S32);
106 MD5_STEP (MD5_I , b, c, d, a, w5_t, MD5C33, MD5S33);
107 MD5_STEP (MD5_I , a, b, c, d, wc_t, MD5C34, MD5S30);
108 MD5_STEP (MD5_I , d, a, b, c, w3_t, MD5C35, MD5S31);
109 MD5_STEP (MD5_I , c, d, a, b, wa_t, MD5C36, MD5S32);
110 MD5_STEP (MD5_I , b, c, d, a, w1_t, MD5C37, MD5S33);
111 MD5_STEP (MD5_I , a, b, c, d, w8_t, MD5C38, MD5S30);
112 MD5_STEP (MD5_I , d, a, b, c, wf_t, MD5C39, MD5S31);
113 MD5_STEP (MD5_I , c, d, a, b, w6_t, MD5C3a, MD5S32);
114 MD5_STEP (MD5_I , b, c, d, a, wd_t, MD5C3b, MD5S33);
115 MD5_STEP (MD5_I , a, b, c, d, w4_t, MD5C3c, MD5S30);
116 MD5_STEP (MD5_I , d, a, b, c, wb_t, MD5C3d, MD5S31);
117 MD5_STEP (MD5_I , c, d, a, b, w2_t, MD5C3e, MD5S32);
118 MD5_STEP (MD5_I , b, c, d, a, w9_t, MD5C3f, MD5S33);
126 static void memcat16 (u32 block0[4], u32 block1[4], u32 block2[4], u32 block3[4], const u32 block_len, const u32 append[4])
136 const int offset_minus_4 = 4 - (block_len & 3);
138 tmp0 = amd_bytealign (append[0], 0, offset_minus_4);
139 tmp1 = amd_bytealign (append[1], append[0], offset_minus_4);
140 tmp2 = amd_bytealign (append[2], append[1], offset_minus_4);
141 tmp3 = amd_bytealign (append[3], append[2], offset_minus_4);
142 tmp4 = amd_bytealign ( 0, append[3], offset_minus_4);
144 const u32 mod = block_len & 3;
159 const int offset_minus_4 = 4 - (block_len & 3);
161 const int selector = (0x76543210 >> (offset_minus_4 * 4)) & 0xffff;
163 tmp0 = __byte_perm ( 0, append[0], selector);
164 tmp1 = __byte_perm (append[0], append[1], selector);
165 tmp2 = __byte_perm (append[1], append[2], selector);
166 tmp3 = __byte_perm (append[2], append[3], selector);
167 tmp4 = __byte_perm (append[3], 0, selector);
171 const u32 div = block_len / 4;
175 case 0: block0[0] |= tmp0;
181 case 1: block0[1] |= tmp0;
187 case 2: block0[2] |= tmp0;
193 case 3: block0[3] |= tmp0;
199 case 4: block1[0] |= tmp0;
205 case 5: block1[1] |= tmp0;
211 case 6: block1[2] |= tmp0;
217 case 7: block1[3] |= tmp0;
223 case 8: block2[0] |= tmp0;
229 case 9: block2[1] |= tmp0;
238 static void memcat16_x80 (u32 block0[4], u32 block1[4], u32 block2[4], u32 block3[4], const u32 block_len, const u32 append[4])
248 const int offset_minus_4 = 4 - (block_len & 3);
250 tmp0 = amd_bytealign (append[0], 0, offset_minus_4);
251 tmp1 = amd_bytealign (append[1], append[0], offset_minus_4);
252 tmp2 = amd_bytealign (append[2], append[1], offset_minus_4);
253 tmp3 = amd_bytealign (append[3], append[2], offset_minus_4);
254 tmp4 = amd_bytealign ( 0x80, append[3], offset_minus_4);
256 const u32 mod = block_len & 3;
271 const int offset_minus_4 = 4 - (block_len & 3);
273 const int selector = (0x76543210 >> (offset_minus_4 * 4)) & 0xffff;
275 tmp0 = __byte_perm ( 0, append[0], selector);
276 tmp1 = __byte_perm (append[0], append[1], selector);
277 tmp2 = __byte_perm (append[1], append[2], selector);
278 tmp3 = __byte_perm (append[2], append[3], selector);
279 tmp4 = __byte_perm (append[3], 0x80, selector);
283 const u32 div = block_len / 4;
287 case 0: block0[0] |= tmp0;
293 case 1: block0[1] |= tmp0;
299 case 2: block0[2] |= tmp0;
305 case 3: block0[3] |= tmp0;
311 case 4: block1[0] |= tmp0;
317 case 5: block1[1] |= tmp0;
323 case 6: block1[2] |= tmp0;
329 case 7: block1[3] |= tmp0;
335 case 8: block2[0] |= tmp0;
341 case 9: block2[1] |= tmp0;
350 static void memcat8 (u32 block0[4], u32 block1[4], u32 block2[4], u32 block3[4], const u32 block_len, const u32 append[2])
358 const int offset_minus_4 = 4 - (block_len & 3);
360 tmp0 = amd_bytealign (append[0], 0, offset_minus_4);
361 tmp1 = amd_bytealign (append[1], append[0], offset_minus_4);
362 tmp2 = amd_bytealign ( 0, append[1], offset_minus_4);
364 const u32 mod = block_len & 3;
377 const int offset_minus_4 = 4 - (block_len & 3);
379 const int selector = (0x76543210 >> (offset_minus_4 * 4)) & 0xffff;
381 tmp0 = __byte_perm ( 0, append[0], selector);
382 tmp1 = __byte_perm (append[0], append[1], selector);
383 tmp2 = __byte_perm (append[1], 0, selector);
387 const u32 div = block_len / 4;
391 case 0: block0[0] |= tmp0;
395 case 1: block0[1] |= tmp0;
399 case 2: block0[2] |= tmp0;
403 case 3: block0[3] |= tmp0;
407 case 4: block1[0] |= tmp0;
411 case 5: block1[1] |= tmp0;
415 case 6: block1[2] |= tmp0;
419 case 7: block1[3] |= tmp0;
423 case 8: block2[0] |= tmp0;
427 case 9: block2[1] |= tmp0;
431 case 10: block2[2] |= tmp0;
435 case 11: block2[3] |= tmp0;
442 static void append_sign (u32 block0[4], u32 block1[4], const u32 block_len)
447 block0[0] = md5apr1_magic0;
448 block0[1] = md5apr1_magic1;
452 block0[0] = block0[0] | md5apr1_magic0 << 8;
453 block0[1] = md5apr1_magic0 >> 24 | md5apr1_magic1 << 8;
454 block0[2] = md5apr1_magic1 >> 24;
458 block0[0] = block0[0] | md5apr1_magic0 << 16;
459 block0[1] = md5apr1_magic0 >> 16 | md5apr1_magic1 << 16;
460 block0[2] = md5apr1_magic1 >> 16;
464 block0[0] = block0[0] | md5apr1_magic0 << 24;
465 block0[1] = md5apr1_magic0 >> 8 | md5apr1_magic1 << 24;
466 block0[2] = md5apr1_magic1 >> 8;
470 block0[1] = md5apr1_magic0;
471 block0[2] = md5apr1_magic1;
475 block0[1] = block0[1] | md5apr1_magic0 << 8;
476 block0[2] = md5apr1_magic0 >> 24 | md5apr1_magic1 << 8;
477 block0[3] = md5apr1_magic1 >> 24;
481 block0[1] = block0[1] | md5apr1_magic0 << 16;
482 block0[2] = md5apr1_magic0 >> 16 | md5apr1_magic1 << 16;
483 block0[3] = md5apr1_magic1 >> 16;
487 block0[1] = block0[1] | md5apr1_magic0 << 24;
488 block0[2] = md5apr1_magic0 >> 8 | md5apr1_magic1 << 24;
489 block0[3] = md5apr1_magic1 >> 8;
493 block0[2] = md5apr1_magic0;
494 block0[3] = md5apr1_magic1;
498 block0[2] = block0[2] | md5apr1_magic0 << 8;
499 block0[3] = md5apr1_magic0 >> 24 | md5apr1_magic1 << 8;
500 block1[0] = md5apr1_magic1 >> 24;
504 block0[2] = block0[2] | md5apr1_magic0 << 16;
505 block0[3] = md5apr1_magic0 >> 16 | md5apr1_magic1 << 16;
506 block1[0] = md5apr1_magic1 >> 16;
510 block0[2] = block0[2] | md5apr1_magic0 << 24;
511 block0[3] = md5apr1_magic0 >> 8 | md5apr1_magic1 << 24;
512 block1[0] = md5apr1_magic1 >> 8;
516 block0[3] = md5apr1_magic0;
517 block1[0] = md5apr1_magic1;
521 block0[3] = block0[3] | md5apr1_magic0 << 8;
522 block1[0] = md5apr1_magic0 >> 24 | md5apr1_magic1 << 8;
523 block1[1] = md5apr1_magic1 >> 24;
527 block0[3] = block0[3] | md5apr1_magic0 << 16;
528 block1[0] = md5apr1_magic0 >> 16 | md5apr1_magic1 << 16;
529 block1[1] = md5apr1_magic1 >> 16;
533 block0[3] = block0[3] | md5apr1_magic0 << 24;
534 block1[0] = md5apr1_magic0 >> 8 | md5apr1_magic1 << 24;
535 block1[1] = md5apr1_magic1 >> 8;
540 static void append_1st (u32 block0[4], u32 block1[4], u32 block2[4], u32 block3[4], const u32 block_len, const u32 append)
549 block0[0] = block0[0] | append << 8;
553 block0[0] = block0[0] | append << 16;
557 block0[0] = block0[0] | append << 24;
565 block0[1] = block0[1] | append << 8;
569 block0[1] = block0[1] | append << 16;
573 block0[1] = block0[1] | append << 24;
581 block0[2] = block0[2] | append << 8;
585 block0[2] = block0[2] | append << 16;
589 block0[2] = block0[2] | append << 24;
597 block0[3] = block0[3] | append << 8;
601 block0[3] = block0[3] | append << 16;
605 block0[3] = block0[3] | append << 24;
613 block1[0] = block1[0] | append << 8;
617 block1[0] = block1[0] | append << 16;
621 block1[0] = block1[0] | append << 24;
629 block1[1] = block1[1] | append << 8;
633 block1[1] = block1[1] | append << 16;
637 block1[1] = block1[1] | append << 24;
645 block1[2] = block1[2] | append << 8;
649 block1[2] = block1[2] | append << 16;
653 block1[2] = block1[2] | append << 24;
661 block1[3] = block1[3] | append << 8;
665 block1[3] = block1[3] | append << 16;
669 block1[3] = block1[3] | append << 24;
677 block2[0] = block2[0] | append << 8;
681 block2[0] = block2[0] | append << 16;
685 block2[0] = block2[0] | append << 24;
693 block2[1] = block2[1] | append << 8;
697 block2[1] = block2[1] | append << 16;
701 block2[1] = block2[1] | append << 24;
709 block2[2] = block2[2] | append << 8;
713 block2[2] = block2[2] | append << 16;
717 block2[2] = block2[2] | append << 24;
725 block2[3] = block2[3] | append << 8;
729 block2[3] = block2[3] | append << 16;
733 block2[3] = block2[3] | append << 24;
741 block3[0] = block3[0] | append << 8;
745 block3[0] = block3[0] | append << 16;
749 block3[0] = block3[0] | append << 24;
757 block3[1] = block3[1] | append << 8;
761 block3[1] = block3[1] | append << 16;
765 block3[1] = block3[1] | append << 24;
774 __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m01600_init (__global pw_t *pws, __global gpu_rule_t *rules_buf, __global comb_t *combs_buf, __global bf_t *bfs_buf, __global md5crypt_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)
780 const u32 gid = get_global_id (0);
782 if (gid >= gid_max) return;
786 w0[0] = pws[gid].i[0];
787 w0[1] = pws[gid].i[1];
788 w0[2] = pws[gid].i[2];
789 w0[3] = pws[gid].i[3];
791 const u32 pw_len = pws[gid].pw_len;
799 salt_buf[0] = salt_bufs[salt_pos].salt_buf[0];
800 salt_buf[1] = salt_bufs[salt_pos].salt_buf[1];
802 const u32 salt_len = salt_bufs[salt_pos].salt_len;
808 //memcat16 (block0, block1, block2, block3, block_len, w0);
809 //block_len += pw_len;
811 u32 block_len = pw_len;
841 memcat8 (block0, block1, block2, block3, block_len, salt_buf);
843 block_len += salt_len;
845 memcat16 (block0, block1, block2, block3, block_len, w0);
849 append_0x80_4x4 (block0, block1, block2, block3, block_len);
851 block3[2] = block_len * 8;
860 md5_transform (block0, block1, block2, block3, digest);
862 /* The password first, since that is what is most unknown */
863 /* Then our magic string */
864 /* Then the raw salt */
865 /* Then just as many characters of the MD5(pw,salt,pw) */
867 //memcat16 (block0, block1, block2, block3, block_len, w);
868 //block_len += pw_len;
892 append_sign (block0, block1, block_len);
896 memcat8 (block0, block1, block2, block3, block_len, salt_buf);
898 block_len += salt_len;
900 truncate_block (digest, pw_len);
902 memcat16 (block0, block1, block2, block3, block_len, digest);
906 /* Then something really weird... */
908 u32 append = block0[0] & 0xFF;
910 for (u32 j = pw_len; j; j >>= 1)
914 append_1st (block0, block1, block2, block3, block_len, append);
920 append_0x80_4x4 (block0, block1, block2, block3, block_len);
922 block3[2] = block_len * 8;
929 md5_transform (block0, block1, block2, block3, digest);
931 tmps[gid].digest_buf[0] = digest[0];
932 tmps[gid].digest_buf[1] = digest[1];
933 tmps[gid].digest_buf[2] = digest[2];
934 tmps[gid].digest_buf[3] = digest[3];
937 __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m01600_loop (__global pw_t *pws, __global gpu_rule_t *rules_buf, __global comb_t *combs_buf, __global bf_t *bfs_buf, __global md5crypt_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)
943 const u32 gid = get_global_id (0);
945 if (gid >= gid_max) return;
949 w0[0] = pws[gid].i[0];
950 w0[1] = pws[gid].i[1];
951 w0[2] = pws[gid].i[2];
952 w0[3] = pws[gid].i[3];
954 const u32 pw_len = pws[gid].pw_len;
963 append_0x80_1x4 (w0_x80, pw_len);
971 salt_buf[0] = salt_bufs[salt_pos].salt_buf[0];
972 salt_buf[1] = salt_bufs[salt_pos].salt_buf[1];
974 const u32 salt_len = salt_bufs[salt_pos].salt_len;
982 digest[0] = tmps[gid].digest_buf[0];
983 digest[1] = tmps[gid].digest_buf[1];
984 digest[2] = tmps[gid].digest_buf[2];
985 digest[3] = tmps[gid].digest_buf[3];
991 /* and now, just to make sure things don't run too fast */
1023 for (u32 i = 0, j = loop_pos; i < loop_cnt; i++, j++)
1036 const u32 j1 = (j & 1) ? 1 : 0;
1037 const u32 j3 = (j % 3) ? 1 : 0;
1038 const u32 j7 = (j % 7) ? 1 : 0;
1051 memcat8 (block0, block1, block2, block3, block_len, salt_buf);
1053 block_len += salt_len;
1058 memcat16 (block0, block1, block2, block3, block_len, w0);
1060 block_len += pw_len;
1063 memcat16_x80 (block0, block1, block2, block3, block_len, digest);
1069 block0[0] = digest[0];
1070 block0[1] = digest[1];
1071 block0[2] = digest[2];
1072 block0[3] = digest[3];
1078 block1[0] = salt_buf[0];
1079 block1[1] = salt_buf[1];
1081 block_len += salt_len;
1083 memcat16 (block0, block1, block2, block3, block_len, w0);
1085 block_len += pw_len;
1089 block1[0] = salt_buf[0];
1090 block1[1] = salt_buf[1];
1092 block_len += salt_len;
1101 block_len += pw_len;
1104 memcat16 (block0, block1, block2, block3, block_len, w0_x80);
1106 block_len += pw_len;
1109 block3[2] = block_len * 8;
1116 md5_transform (block0, block1, block2, block3, digest);
1119 tmps[gid].digest_buf[0] = digest[0];
1120 tmps[gid].digest_buf[1] = digest[1];
1121 tmps[gid].digest_buf[2] = digest[2];
1122 tmps[gid].digest_buf[3] = digest[3];
1125 __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m01600_comp (__global pw_t *pws, __global gpu_rule_t *rules_buf, __global comb_t *combs_buf, __global bf_t *bfs_buf, __global md5crypt_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)
1131 const u32 gid = get_global_id (0);
1133 if (gid >= gid_max) return;
1135 const u32 lid = get_local_id (0);
1141 const u32 r0 = tmps[gid].digest_buf[DGST_R0];
1142 const u32 r1 = tmps[gid].digest_buf[DGST_R1];
1143 const u32 r2 = tmps[gid].digest_buf[DGST_R2];
1144 const u32 r3 = tmps[gid].digest_buf[DGST_R3];