2 * Author......: Jens Steube <jens.steube@gmail.com>
8 #include "include/constants.h"
9 #include "include/kernel_vendor.h"
24 #include "include/kernel_functions.c"
26 #include "common_nv.c"
29 #define VECT_COMPARE_M "check_multi_vect1_comp4.c"
33 #define VECT_COMPARE_M "check_multi_vect2_comp4.c"
37 #define VECT_COMPARE_M "check_multi_vect4_comp4.c"
40 #define md5crypt_magic 0x00243124
42 __device__ static void md5_transform (const u32x w0[4], const u32x w1[4], const u32x w2[4], const u32x w3[4], u32x digest[4])
68 MD5_STEP (MD5_Fo, a, b, c, d, w0_t, MD5C00, MD5S00);
69 MD5_STEP (MD5_Fo, d, a, b, c, w1_t, MD5C01, MD5S01);
70 MD5_STEP (MD5_Fo, c, d, a, b, w2_t, MD5C02, MD5S02);
71 MD5_STEP (MD5_Fo, b, c, d, a, w3_t, MD5C03, MD5S03);
72 MD5_STEP (MD5_Fo, a, b, c, d, w4_t, MD5C04, MD5S00);
73 MD5_STEP (MD5_Fo, d, a, b, c, w5_t, MD5C05, MD5S01);
74 MD5_STEP (MD5_Fo, c, d, a, b, w6_t, MD5C06, MD5S02);
75 MD5_STEP (MD5_Fo, b, c, d, a, w7_t, MD5C07, MD5S03);
76 MD5_STEP (MD5_Fo, a, b, c, d, w8_t, MD5C08, MD5S00);
77 MD5_STEP (MD5_Fo, d, a, b, c, w9_t, MD5C09, MD5S01);
78 MD5_STEP (MD5_Fo, c, d, a, b, wa_t, MD5C0a, MD5S02);
79 MD5_STEP (MD5_Fo, b, c, d, a, wb_t, MD5C0b, MD5S03);
80 MD5_STEP (MD5_Fo, a, b, c, d, wc_t, MD5C0c, MD5S00);
81 MD5_STEP (MD5_Fo, d, a, b, c, wd_t, MD5C0d, MD5S01);
82 MD5_STEP (MD5_Fo, c, d, a, b, we_t, MD5C0e, MD5S02);
83 MD5_STEP (MD5_Fo, b, c, d, a, wf_t, MD5C0f, MD5S03);
85 MD5_STEP (MD5_Go, a, b, c, d, w1_t, MD5C10, MD5S10);
86 MD5_STEP (MD5_Go, d, a, b, c, w6_t, MD5C11, MD5S11);
87 MD5_STEP (MD5_Go, c, d, a, b, wb_t, MD5C12, MD5S12);
88 MD5_STEP (MD5_Go, b, c, d, a, w0_t, MD5C13, MD5S13);
89 MD5_STEP (MD5_Go, a, b, c, d, w5_t, MD5C14, MD5S10);
90 MD5_STEP (MD5_Go, d, a, b, c, wa_t, MD5C15, MD5S11);
91 MD5_STEP (MD5_Go, c, d, a, b, wf_t, MD5C16, MD5S12);
92 MD5_STEP (MD5_Go, b, c, d, a, w4_t, MD5C17, MD5S13);
93 MD5_STEP (MD5_Go, a, b, c, d, w9_t, MD5C18, MD5S10);
94 MD5_STEP (MD5_Go, d, a, b, c, we_t, MD5C19, MD5S11);
95 MD5_STEP (MD5_Go, c, d, a, b, w3_t, MD5C1a, MD5S12);
96 MD5_STEP (MD5_Go, b, c, d, a, w8_t, MD5C1b, MD5S13);
97 MD5_STEP (MD5_Go, a, b, c, d, wd_t, MD5C1c, MD5S10);
98 MD5_STEP (MD5_Go, d, a, b, c, w2_t, MD5C1d, MD5S11);
99 MD5_STEP (MD5_Go, c, d, a, b, w7_t, MD5C1e, MD5S12);
100 MD5_STEP (MD5_Go, b, c, d, a, wc_t, MD5C1f, MD5S13);
102 MD5_STEP (MD5_H1, a, b, c, d, w5_t, MD5C20, MD5S20);
103 MD5_STEP (MD5_H2, d, a, b, c, w8_t, MD5C21, MD5S21);
104 MD5_STEP (MD5_H1, c, d, a, b, wb_t, MD5C22, MD5S22);
105 MD5_STEP (MD5_H2, b, c, d, a, we_t, MD5C23, MD5S23);
106 MD5_STEP (MD5_H1, a, b, c, d, w1_t, MD5C24, MD5S20);
107 MD5_STEP (MD5_H2, d, a, b, c, w4_t, MD5C25, MD5S21);
108 MD5_STEP (MD5_H1, c, d, a, b, w7_t, MD5C26, MD5S22);
109 MD5_STEP (MD5_H2, b, c, d, a, wa_t, MD5C27, MD5S23);
110 MD5_STEP (MD5_H1, a, b, c, d, wd_t, MD5C28, MD5S20);
111 MD5_STEP (MD5_H2, d, a, b, c, w0_t, MD5C29, MD5S21);
112 MD5_STEP (MD5_H1, c, d, a, b, w3_t, MD5C2a, MD5S22);
113 MD5_STEP (MD5_H2, b, c, d, a, w6_t, MD5C2b, MD5S23);
114 MD5_STEP (MD5_H1, a, b, c, d, w9_t, MD5C2c, MD5S20);
115 MD5_STEP (MD5_H2, d, a, b, c, wc_t, MD5C2d, MD5S21);
116 MD5_STEP (MD5_H1, c, d, a, b, wf_t, MD5C2e, MD5S22);
117 MD5_STEP (MD5_H2, b, c, d, a, w2_t, MD5C2f, MD5S23);
119 MD5_STEP (MD5_I , a, b, c, d, w0_t, MD5C30, MD5S30);
120 MD5_STEP (MD5_I , d, a, b, c, w7_t, MD5C31, MD5S31);
121 MD5_STEP (MD5_I , c, d, a, b, we_t, MD5C32, MD5S32);
122 MD5_STEP (MD5_I , b, c, d, a, w5_t, MD5C33, MD5S33);
123 MD5_STEP (MD5_I , a, b, c, d, wc_t, MD5C34, MD5S30);
124 MD5_STEP (MD5_I , d, a, b, c, w3_t, MD5C35, MD5S31);
125 MD5_STEP (MD5_I , c, d, a, b, wa_t, MD5C36, MD5S32);
126 MD5_STEP (MD5_I , b, c, d, a, w1_t, MD5C37, MD5S33);
127 MD5_STEP (MD5_I , a, b, c, d, w8_t, MD5C38, MD5S30);
128 MD5_STEP (MD5_I , d, a, b, c, wf_t, MD5C39, MD5S31);
129 MD5_STEP (MD5_I , c, d, a, b, w6_t, MD5C3a, MD5S32);
130 MD5_STEP (MD5_I , b, c, d, a, wd_t, MD5C3b, MD5S33);
131 MD5_STEP (MD5_I , a, b, c, d, w4_t, MD5C3c, MD5S30);
132 MD5_STEP (MD5_I , d, a, b, c, wb_t, MD5C3d, MD5S31);
133 MD5_STEP (MD5_I , c, d, a, b, w2_t, MD5C3e, MD5S32);
134 MD5_STEP (MD5_I , b, c, d, a, w9_t, MD5C3f, MD5S33);
142 __device__ static void memcat16 (u32x block0[4], u32x block1[4], u32x block2[4], u32x block3[4], const u32 block_len, const u32x append[4])
150 #if __CUDA_ARCH__ >= 200
152 const int offset_minus_4 = 4 - (block_len & 3);
154 const int selector = (0x76543210 >> (offset_minus_4 * 4)) & 0xffff;
156 tmp0 = __byte_perm ( 0, append[0], selector);
157 tmp1 = __byte_perm (append[0], append[1], selector);
158 tmp2 = __byte_perm (append[1], append[2], selector);
159 tmp3 = __byte_perm (append[2], append[3], selector);
160 tmp4 = __byte_perm (append[3], 0, selector);
164 const u32 mod = block_len & 3;
168 case 0: tmp0 = append[0];
174 case 1: tmp0 = append[0] << 8;
175 tmp1 = append[0] >> 24 | append[1] << 8;
176 tmp2 = append[1] >> 24 | append[2] << 8;
177 tmp3 = append[2] >> 24 | append[3] << 8;
178 tmp4 = append[3] >> 24;
180 case 2: tmp0 = append[0] << 16;
181 tmp1 = append[0] >> 16 | append[1] << 16;
182 tmp2 = append[1] >> 16 | append[2] << 16;
183 tmp3 = append[2] >> 16 | append[3] << 16;
184 tmp4 = append[3] >> 16;
186 case 3: tmp0 = append[0] << 24;
187 tmp1 = append[0] >> 8 | append[1] << 24;
188 tmp2 = append[1] >> 8 | append[2] << 24;
189 tmp3 = append[2] >> 8 | append[3] << 24;
190 tmp4 = append[3] >> 8;
196 const u32 div = block_len / 4;
200 case 0: block0[0] |= tmp0;
206 case 1: block0[1] |= tmp0;
212 case 2: block0[2] |= tmp0;
218 case 3: block0[3] |= tmp0;
224 case 4: block1[0] |= tmp0;
230 case 5: block1[1] |= tmp0;
236 case 6: block1[2] |= tmp0;
242 case 7: block1[3] |= tmp0;
248 case 8: block2[0] |= tmp0;
254 case 9: block2[1] |= tmp0;
265 __device__ static void memcat16_x80 (u32x block0[4], u32x block1[4], u32x block2[4], u32x block3[4], const u32 block_len, const u32x append[4])
273 #if __CUDA_ARCH__ >= 200
275 const int offset_minus_4 = 4 - (block_len & 3);
277 const int selector = (0x76543210 >> (offset_minus_4 * 4)) & 0xffff;
279 tmp0 = __byte_perm ( 0, append[0], selector);
280 tmp1 = __byte_perm (append[0], append[1], selector);
281 tmp2 = __byte_perm (append[1], append[2], selector);
282 tmp3 = __byte_perm (append[2], append[3], selector);
283 tmp4 = __byte_perm (append[3], 0x80, selector);
287 const u32 mod = block_len & 3;
291 case 0: tmp0 = append[0];
297 case 1: tmp0 = append[0] << 8;
298 tmp1 = append[0] >> 24 | append[1] << 8;
299 tmp2 = append[1] >> 24 | append[2] << 8;
300 tmp3 = append[2] >> 24 | append[3] << 8;
301 tmp4 = append[3] >> 24;
303 case 2: tmp0 = append[0] << 16;
304 tmp1 = append[0] >> 16 | append[1] << 16;
305 tmp2 = append[1] >> 16 | append[2] << 16;
306 tmp3 = append[2] >> 16 | append[3] << 16;
307 tmp4 = append[3] >> 16;
309 case 3: tmp0 = append[0] << 24;
310 tmp1 = append[0] >> 8 | append[1] << 24;
311 tmp2 = append[1] >> 8 | append[2] << 24;
312 tmp3 = append[2] >> 8 | append[3] << 24;
313 tmp4 = append[3] >> 8;
319 const u32 div = block_len / 4;
323 case 0: block0[0] |= tmp0;
329 case 1: block0[1] |= tmp0;
335 case 2: block0[2] |= tmp0;
341 case 3: block0[3] |= tmp0;
347 case 4: block1[0] |= tmp0;
353 case 5: block1[1] |= tmp0;
359 case 6: block1[2] |= tmp0;
365 case 7: block1[3] |= tmp0;
371 case 8: block2[0] |= tmp0;
377 case 9: block2[1] |= tmp0;
388 __device__ static void memcat8 (u32x block0[4], u32x block1[4], u32x block2[4], u32x block3[4], const u32 block_len, const u32 append[2])
394 #if __CUDA_ARCH__ >= 200
396 const int offset_minus_4 = 4 - (block_len & 3);
398 const int selector = (0x76543210 >> (offset_minus_4 * 4)) & 0xffff;
400 tmp0 = __byte_perm ( 0, append[0], selector);
401 tmp1 = __byte_perm (append[0], append[1], selector);
402 tmp2 = __byte_perm (append[1], 0, selector);
406 const u32 mod = block_len & 3;
410 case 0: tmp0 = append[0];
414 case 1: tmp0 = append[0] << 8;
415 tmp1 = append[0] >> 24 | append[1] << 8;
416 tmp2 = append[1] >> 24;
418 case 2: tmp0 = append[0] << 16;
419 tmp1 = append[0] >> 16 | append[1] << 16;
420 tmp2 = append[1] >> 16;
422 case 3: tmp0 = append[0] << 24;
423 tmp1 = append[0] >> 8 | append[1] << 24;
424 tmp2 = append[1] >> 8;
430 const u32 div = block_len / 4;
434 case 0: block0[0] |= tmp0;
438 case 1: block0[1] |= tmp0;
442 case 2: block0[2] |= tmp0;
446 case 3: block0[3] |= tmp0;
450 case 4: block1[0] |= tmp0;
454 case 5: block1[1] |= tmp0;
458 case 6: block1[2] |= tmp0;
462 case 7: block1[3] |= tmp0;
466 case 8: block2[0] |= tmp0;
470 case 9: block2[1] |= tmp0;
474 case 10: block2[2] |= tmp0;
478 case 11: block2[3] |= tmp0;
487 __device__ static void append_sign (u32x block0[4], u32x block1[4], const u32 block_len)
492 block0[0] = md5crypt_magic;
496 block0[0] = block0[0] | md5crypt_magic << 8;
497 block0[1] = md5crypt_magic >> 24;
501 block0[0] = block0[0] | md5crypt_magic << 16;
502 block0[1] = md5crypt_magic >> 16;
506 block0[0] = block0[0] | md5crypt_magic << 24;
507 block0[1] = md5crypt_magic >> 8;
511 block0[1] = md5crypt_magic;
515 block0[1] = block0[1] | md5crypt_magic << 8;
516 block0[2] = md5crypt_magic >> 24;
520 block0[1] = block0[1] | md5crypt_magic << 16;
521 block0[2] = md5crypt_magic >> 16;
525 block0[1] = block0[1] | md5crypt_magic << 24;
526 block0[2] = md5crypt_magic >> 8;
530 block0[2] = md5crypt_magic;
534 block0[2] = block0[2] | md5crypt_magic << 8;
535 block0[3] = md5crypt_magic >> 24;
539 block0[2] = block0[2] | md5crypt_magic << 16;
540 block0[3] = md5crypt_magic >> 16;
544 block0[2] = block0[2] | md5crypt_magic << 24;
545 block0[3] = md5crypt_magic >> 8;
549 block0[3] = md5crypt_magic;
553 block0[3] = block0[3] | md5crypt_magic << 8;
554 block1[0] = md5crypt_magic >> 24;
558 block0[3] = block0[3] | md5crypt_magic << 16;
559 block1[0] = md5crypt_magic >> 16;
563 block0[3] = block0[3] | md5crypt_magic << 24;
564 block1[0] = md5crypt_magic >> 8;
569 __device__ static void append_1st (u32x block0[4], u32x block1[4], u32x block2[4], u32x block3[4], const u32 block_len, const u32x append)
578 block0[0] = block0[0] | append << 8;
582 block0[0] = block0[0] | append << 16;
586 block0[0] = block0[0] | append << 24;
594 block0[1] = block0[1] | append << 8;
598 block0[1] = block0[1] | append << 16;
602 block0[1] = block0[1] | append << 24;
610 block0[2] = block0[2] | append << 8;
614 block0[2] = block0[2] | append << 16;
618 block0[2] = block0[2] | append << 24;
626 block0[3] = block0[3] | append << 8;
630 block0[3] = block0[3] | append << 16;
634 block0[3] = block0[3] | append << 24;
642 block1[0] = block1[0] | append << 8;
646 block1[0] = block1[0] | append << 16;
650 block1[0] = block1[0] | append << 24;
658 block1[1] = block1[1] | append << 8;
662 block1[1] = block1[1] | append << 16;
666 block1[1] = block1[1] | append << 24;
674 block1[2] = block1[2] | append << 8;
678 block1[2] = block1[2] | append << 16;
682 block1[2] = block1[2] | append << 24;
690 block1[3] = block1[3] | append << 8;
694 block1[3] = block1[3] | append << 16;
698 block1[3] = block1[3] | append << 24;
706 block2[0] = block2[0] | append << 8;
710 block2[0] = block2[0] | append << 16;
714 block2[0] = block2[0] | append << 24;
722 block2[1] = block2[1] | append << 8;
726 block2[1] = block2[1] | append << 16;
730 block2[1] = block2[1] | append << 24;
738 block2[2] = block2[2] | append << 8;
742 block2[2] = block2[2] | append << 16;
746 block2[2] = block2[2] | append << 24;
754 block2[3] = block2[3] | append << 8;
758 block2[3] = block2[3] | append << 16;
762 block2[3] = block2[3] | append << 24;
770 block3[0] = block3[0] | append << 8;
774 block3[0] = block3[0] | append << 16;
778 block3[0] = block3[0] | append << 24;
786 block3[1] = block3[1] | append << 8;
790 block3[1] = block3[1] | append << 16;
794 block3[1] = block3[1] | append << 24;
803 extern "C" __global__ void __launch_bounds__ (256, 1) m00500_init (const pw_t *pws, const gpu_rule_t *rules_buf, const comb_t *combs_buf, const bf_t *bfs_buf, md5crypt_tmp_t *tmps, void *hooks, const u32 *bitmaps_buf_s1_a, const u32 *bitmaps_buf_s1_b, const u32 *bitmaps_buf_s1_c, const u32 *bitmaps_buf_s1_d, const u32 *bitmaps_buf_s2_a, const u32 *bitmaps_buf_s2_b, const u32 *bitmaps_buf_s2_c, const u32 *bitmaps_buf_s2_d, plain_t *plains_buf, const digest_t *digests_buf, u32 *hashes_shown, const salt_t *salt_bufs, const void *esalt_bufs, u32 *d_return_buf, 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)
805 const u32 gid = (blockIdx.x * blockDim.x) + threadIdx.x;
807 if (gid >= gid_max) return;
811 w0[0] = pws[gid].i[0];
812 w0[1] = pws[gid].i[1];
813 w0[2] = pws[gid].i[2];
814 w0[3] = pws[gid].i[3];
816 const u32 pw_len = pws[gid].pw_len;
824 salt_buf[0] = salt_bufs[salt_pos].salt_buf[0];
825 salt_buf[1] = salt_bufs[salt_pos].salt_buf[1];
827 const u32 salt_len = salt_bufs[salt_pos].salt_len;
833 //memcat16 (block0, block1, block2, block3, block_len, w0);
834 //block_len += pw_len;
836 u32 block_len = pw_len;
866 memcat8 (block0, block1, block2, block3, block_len, salt_buf);
868 block_len += salt_len;
870 memcat16 (block0, block1, block2, block3, block_len, w0);
874 append_0x80_4 (block0, block1, block2, block3, block_len);
876 block3[2] = block_len * 8;
885 md5_transform (block0, block1, block2, block3, digest);
887 /* The password first, since that is what is most unknown */
888 /* Then our magic string */
889 /* Then the raw salt */
890 /* Then just as many characters of the MD5(pw,salt,pw) */
892 //memcat16 (block0, block1, block2, block3, block_len, w);
893 //block_len += pw_len;
917 append_sign (block0, block1, block_len);
921 memcat8 (block0, block1, block2, block3, block_len, salt_buf);
923 block_len += salt_len;
925 truncate_block (digest, pw_len);
927 memcat16 (block0, block1, block2, block3, block_len, digest);
931 /* Then something really weird... */
933 u32x append = block0[0] & 0xFF;
935 for (u32 j = pw_len; j; j >>= 1)
939 append_1st (block0, block1, block2, block3, block_len, append);
945 append_0x80_4 (block0, block1, block2, block3, block_len);
947 block3[2] = block_len * 8;
954 md5_transform (block0, block1, block2, block3, digest);
956 tmps[gid].digest_buf[0] = digest[0];
957 tmps[gid].digest_buf[1] = digest[1];
958 tmps[gid].digest_buf[2] = digest[2];
959 tmps[gid].digest_buf[3] = digest[3];
962 extern "C" __global__ void __launch_bounds__ (256, 1) m00500_loop (const pw_t *pws, const gpu_rule_t *rules_buf, const comb_t *combs_buf, const bf_t *bfs_buf, md5crypt_tmp_t *tmps, void *hooks, const u32 *bitmaps_buf_s1_a, const u32 *bitmaps_buf_s1_b, const u32 *bitmaps_buf_s1_c, const u32 *bitmaps_buf_s1_d, const u32 *bitmaps_buf_s2_a, const u32 *bitmaps_buf_s2_b, const u32 *bitmaps_buf_s2_c, const u32 *bitmaps_buf_s2_d, plain_t *plains_buf, const digest_t *digests_buf, u32 *hashes_shown, const salt_t *salt_bufs, const void *esalt_bufs, u32 *d_return_buf, 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)
968 const u32 gid = (blockIdx.x * blockDim.x) + threadIdx.x;
970 if (gid >= gid_max) return;
974 w0[0] = pws[gid].i[0];
975 w0[1] = pws[gid].i[1];
976 w0[2] = pws[gid].i[2];
977 w0[3] = pws[gid].i[3];
979 const u32 pw_len = pws[gid].pw_len;
988 append_0x80_1 (w0_x80, pw_len);
996 salt_buf[0] = salt_bufs[salt_pos].salt_buf[0];
997 salt_buf[1] = salt_bufs[salt_pos].salt_buf[1];
999 const u32 salt_len = salt_bufs[salt_pos].salt_len;
1007 digest[0] = tmps[gid].digest_buf[0];
1008 digest[1] = tmps[gid].digest_buf[1];
1009 digest[2] = tmps[gid].digest_buf[2];
1010 digest[3] = tmps[gid].digest_buf[3];
1016 /* and now, just to make sure things don't run too fast */
1048 for (u32 i = 0, j = loop_pos; i < loop_cnt; i++, j++)
1061 const u32 j1 = (j & 1) ? 1 : 0;
1062 const u32 j3 = (j % 3) ? 1 : 0;
1063 const u32 j7 = (j % 7) ? 1 : 0;
1076 memcat8 (block0, block1, block2, block3, block_len, salt_buf);
1078 block_len += salt_len;
1083 memcat16 (block0, block1, block2, block3, block_len, w0);
1085 block_len += pw_len;
1088 memcat16_x80 (block0, block1, block2, block3, block_len, digest);
1094 block0[0] = digest[0];
1095 block0[1] = digest[1];
1096 block0[2] = digest[2];
1097 block0[3] = digest[3];
1103 block1[0] = salt_buf[0];
1104 block1[1] = salt_buf[1];
1106 block_len += salt_len;
1108 memcat16 (block0, block1, block2, block3, block_len, w0);
1110 block_len += pw_len;
1114 block1[0] = salt_buf[0];
1115 block1[1] = salt_buf[1];
1117 block_len += salt_len;
1126 block_len += pw_len;
1129 memcat16 (block0, block1, block2, block3, block_len, w0_x80);
1131 block_len += pw_len;
1134 block3[2] = block_len * 8;
1141 md5_transform (block0, block1, block2, block3, digest);
1144 tmps[gid].digest_buf[0] = digest[0];
1145 tmps[gid].digest_buf[1] = digest[1];
1146 tmps[gid].digest_buf[2] = digest[2];
1147 tmps[gid].digest_buf[3] = digest[3];
1150 extern "C" __global__ void __launch_bounds__ (256, 1) m00500_comp (const pw_t *pws, const gpu_rule_t *rules_buf, const comb_t *combs_buf, const bf_t *bfs_buf, md5crypt_tmp_t *tmps, void *hooks, const u32 *bitmaps_buf_s1_a, const u32 *bitmaps_buf_s1_b, const u32 *bitmaps_buf_s1_c, const u32 *bitmaps_buf_s1_d, const u32 *bitmaps_buf_s2_a, const u32 *bitmaps_buf_s2_b, const u32 *bitmaps_buf_s2_c, const u32 *bitmaps_buf_s2_d, plain_t *plains_buf, const digest_t *digests_buf, u32 *hashes_shown, const salt_t *salt_bufs, const void *esalt_bufs, u32 *d_return_buf, 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)
1156 const u32 gid = (blockIdx.x * blockDim.x) + threadIdx.x;
1158 if (gid >= gid_max) return;
1160 const u32 lid = threadIdx.x;
1166 const u32x r0 = tmps[gid].digest_buf[DGST_R0];
1167 const u32x r1 = tmps[gid].digest_buf[DGST_R1];
1168 const u32x r2 = tmps[gid].digest_buf[DGST_R2];
1169 const u32x r3 = tmps[gid].digest_buf[DGST_R3];
1173 #include VECT_COMPARE_M