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"
36 #define md5apr1_magic0 0x72706124
37 #define md5apr1_magic1 0x00002431
39 __device__ static void md5_transform (const u32x w0[4], const u32x w1[4], const u32x w2[4], const u32x w3[4], u32x digest[4])
65 MD5_STEP (MD5_Fo, a, b, c, d, w0_t, MD5C00, MD5S00);
66 MD5_STEP (MD5_Fo, d, a, b, c, w1_t, MD5C01, MD5S01);
67 MD5_STEP (MD5_Fo, c, d, a, b, w2_t, MD5C02, MD5S02);
68 MD5_STEP (MD5_Fo, b, c, d, a, w3_t, MD5C03, MD5S03);
69 MD5_STEP (MD5_Fo, a, b, c, d, w4_t, MD5C04, MD5S00);
70 MD5_STEP (MD5_Fo, d, a, b, c, w5_t, MD5C05, MD5S01);
71 MD5_STEP (MD5_Fo, c, d, a, b, w6_t, MD5C06, MD5S02);
72 MD5_STEP (MD5_Fo, b, c, d, a, w7_t, MD5C07, MD5S03);
73 MD5_STEP (MD5_Fo, a, b, c, d, w8_t, MD5C08, MD5S00);
74 MD5_STEP (MD5_Fo, d, a, b, c, w9_t, MD5C09, MD5S01);
75 MD5_STEP (MD5_Fo, c, d, a, b, wa_t, MD5C0a, MD5S02);
76 MD5_STEP (MD5_Fo, b, c, d, a, wb_t, MD5C0b, MD5S03);
77 MD5_STEP (MD5_Fo, a, b, c, d, wc_t, MD5C0c, MD5S00);
78 MD5_STEP (MD5_Fo, d, a, b, c, wd_t, MD5C0d, MD5S01);
79 MD5_STEP (MD5_Fo, c, d, a, b, we_t, MD5C0e, MD5S02);
80 MD5_STEP (MD5_Fo, b, c, d, a, wf_t, MD5C0f, MD5S03);
82 MD5_STEP (MD5_Go, a, b, c, d, w1_t, MD5C10, MD5S10);
83 MD5_STEP (MD5_Go, d, a, b, c, w6_t, MD5C11, MD5S11);
84 MD5_STEP (MD5_Go, c, d, a, b, wb_t, MD5C12, MD5S12);
85 MD5_STEP (MD5_Go, b, c, d, a, w0_t, MD5C13, MD5S13);
86 MD5_STEP (MD5_Go, a, b, c, d, w5_t, MD5C14, MD5S10);
87 MD5_STEP (MD5_Go, d, a, b, c, wa_t, MD5C15, MD5S11);
88 MD5_STEP (MD5_Go, c, d, a, b, wf_t, MD5C16, MD5S12);
89 MD5_STEP (MD5_Go, b, c, d, a, w4_t, MD5C17, MD5S13);
90 MD5_STEP (MD5_Go, a, b, c, d, w9_t, MD5C18, MD5S10);
91 MD5_STEP (MD5_Go, d, a, b, c, we_t, MD5C19, MD5S11);
92 MD5_STEP (MD5_Go, c, d, a, b, w3_t, MD5C1a, MD5S12);
93 MD5_STEP (MD5_Go, b, c, d, a, w8_t, MD5C1b, MD5S13);
94 MD5_STEP (MD5_Go, a, b, c, d, wd_t, MD5C1c, MD5S10);
95 MD5_STEP (MD5_Go, d, a, b, c, w2_t, MD5C1d, MD5S11);
96 MD5_STEP (MD5_Go, c, d, a, b, w7_t, MD5C1e, MD5S12);
97 MD5_STEP (MD5_Go, b, c, d, a, wc_t, MD5C1f, MD5S13);
99 MD5_STEP (MD5_H1, a, b, c, d, w5_t, MD5C20, MD5S20);
100 MD5_STEP (MD5_H2, d, a, b, c, w8_t, MD5C21, MD5S21);
101 MD5_STEP (MD5_H1, c, d, a, b, wb_t, MD5C22, MD5S22);
102 MD5_STEP (MD5_H2, b, c, d, a, we_t, MD5C23, MD5S23);
103 MD5_STEP (MD5_H1, a, b, c, d, w1_t, MD5C24, MD5S20);
104 MD5_STEP (MD5_H2, d, a, b, c, w4_t, MD5C25, MD5S21);
105 MD5_STEP (MD5_H1, c, d, a, b, w7_t, MD5C26, MD5S22);
106 MD5_STEP (MD5_H2, b, c, d, a, wa_t, MD5C27, MD5S23);
107 MD5_STEP (MD5_H1, a, b, c, d, wd_t, MD5C28, MD5S20);
108 MD5_STEP (MD5_H2, d, a, b, c, w0_t, MD5C29, MD5S21);
109 MD5_STEP (MD5_H1, c, d, a, b, w3_t, MD5C2a, MD5S22);
110 MD5_STEP (MD5_H2, b, c, d, a, w6_t, MD5C2b, MD5S23);
111 MD5_STEP (MD5_H1, a, b, c, d, w9_t, MD5C2c, MD5S20);
112 MD5_STEP (MD5_H2, d, a, b, c, wc_t, MD5C2d, MD5S21);
113 MD5_STEP (MD5_H1, c, d, a, b, wf_t, MD5C2e, MD5S22);
114 MD5_STEP (MD5_H2, b, c, d, a, w2_t, MD5C2f, MD5S23);
116 MD5_STEP (MD5_I , a, b, c, d, w0_t, MD5C30, MD5S30);
117 MD5_STEP (MD5_I , d, a, b, c, w7_t, MD5C31, MD5S31);
118 MD5_STEP (MD5_I , c, d, a, b, we_t, MD5C32, MD5S32);
119 MD5_STEP (MD5_I , b, c, d, a, w5_t, MD5C33, MD5S33);
120 MD5_STEP (MD5_I , a, b, c, d, wc_t, MD5C34, MD5S30);
121 MD5_STEP (MD5_I , d, a, b, c, w3_t, MD5C35, MD5S31);
122 MD5_STEP (MD5_I , c, d, a, b, wa_t, MD5C36, MD5S32);
123 MD5_STEP (MD5_I , b, c, d, a, w1_t, MD5C37, MD5S33);
124 MD5_STEP (MD5_I , a, b, c, d, w8_t, MD5C38, MD5S30);
125 MD5_STEP (MD5_I , d, a, b, c, wf_t, MD5C39, MD5S31);
126 MD5_STEP (MD5_I , c, d, a, b, w6_t, MD5C3a, MD5S32);
127 MD5_STEP (MD5_I , b, c, d, a, wd_t, MD5C3b, MD5S33);
128 MD5_STEP (MD5_I , a, b, c, d, w4_t, MD5C3c, MD5S30);
129 MD5_STEP (MD5_I , d, a, b, c, wb_t, MD5C3d, MD5S31);
130 MD5_STEP (MD5_I , c, d, a, b, w2_t, MD5C3e, MD5S32);
131 MD5_STEP (MD5_I , b, c, d, a, w9_t, MD5C3f, MD5S33);
139 __device__ static void memcat16 (u32x block0[4], u32x block1[4], u32x block2[4], u32x block3[4], const u32 block_len, const u32x append[4])
147 #if __CUDA_ARCH__ >= 200
149 const int offset_minus_4 = 4 - (block_len & 3);
151 const int selector = (0x76543210 >> (offset_minus_4 * 4)) & 0xffff;
153 tmp0 = __byte_perm ( 0, append[0], selector);
154 tmp1 = __byte_perm (append[0], append[1], selector);
155 tmp2 = __byte_perm (append[1], append[2], selector);
156 tmp3 = __byte_perm (append[2], append[3], selector);
157 tmp4 = __byte_perm (append[3], 0, selector);
161 const u32 mod = block_len & 3;
165 case 0: tmp0 = append[0];
171 case 1: tmp0 = append[0] << 8;
172 tmp1 = append[0] >> 24 | append[1] << 8;
173 tmp2 = append[1] >> 24 | append[2] << 8;
174 tmp3 = append[2] >> 24 | append[3] << 8;
175 tmp4 = append[3] >> 24;
177 case 2: tmp0 = append[0] << 16;
178 tmp1 = append[0] >> 16 | append[1] << 16;
179 tmp2 = append[1] >> 16 | append[2] << 16;
180 tmp3 = append[2] >> 16 | append[3] << 16;
181 tmp4 = append[3] >> 16;
183 case 3: tmp0 = append[0] << 24;
184 tmp1 = append[0] >> 8 | append[1] << 24;
185 tmp2 = append[1] >> 8 | append[2] << 24;
186 tmp3 = append[2] >> 8 | append[3] << 24;
187 tmp4 = append[3] >> 8;
193 const u32 div = block_len / 4;
197 case 0: block0[0] |= tmp0;
203 case 1: block0[1] |= tmp0;
209 case 2: block0[2] |= tmp0;
215 case 3: block0[3] |= tmp0;
221 case 4: block1[0] |= tmp0;
227 case 5: block1[1] |= tmp0;
233 case 6: block1[2] |= tmp0;
239 case 7: block1[3] |= tmp0;
245 case 8: block2[0] |= tmp0;
251 case 9: block2[1] |= tmp0;
262 __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])
270 #if __CUDA_ARCH__ >= 200
272 const int offset_minus_4 = 4 - (block_len & 3);
274 const int selector = (0x76543210 >> (offset_minus_4 * 4)) & 0xffff;
276 tmp0 = __byte_perm ( 0, append[0], selector);
277 tmp1 = __byte_perm (append[0], append[1], selector);
278 tmp2 = __byte_perm (append[1], append[2], selector);
279 tmp3 = __byte_perm (append[2], append[3], selector);
280 tmp4 = __byte_perm (append[3], 0x80, selector);
284 const u32 mod = block_len & 3;
288 case 0: tmp0 = append[0];
294 case 1: tmp0 = append[0] << 8;
295 tmp1 = append[0] >> 24 | append[1] << 8;
296 tmp2 = append[1] >> 24 | append[2] << 8;
297 tmp3 = append[2] >> 24 | append[3] << 8;
298 tmp4 = append[3] >> 24;
300 case 2: tmp0 = append[0] << 16;
301 tmp1 = append[0] >> 16 | append[1] << 16;
302 tmp2 = append[1] >> 16 | append[2] << 16;
303 tmp3 = append[2] >> 16 | append[3] << 16;
304 tmp4 = append[3] >> 16;
306 case 3: tmp0 = append[0] << 24;
307 tmp1 = append[0] >> 8 | append[1] << 24;
308 tmp2 = append[1] >> 8 | append[2] << 24;
309 tmp3 = append[2] >> 8 | append[3] << 24;
310 tmp4 = append[3] >> 8;
316 const u32 div = block_len / 4;
320 case 0: block0[0] |= tmp0;
326 case 1: block0[1] |= tmp0;
332 case 2: block0[2] |= tmp0;
338 case 3: block0[3] |= tmp0;
344 case 4: block1[0] |= tmp0;
350 case 5: block1[1] |= tmp0;
356 case 6: block1[2] |= tmp0;
362 case 7: block1[3] |= tmp0;
368 case 8: block2[0] |= tmp0;
374 case 9: block2[1] |= tmp0;
385 __device__ static void memcat8 (u32x block0[4], u32x block1[4], u32x block2[4], u32x block3[4], const u32 block_len, const u32 append[2])
391 #if __CUDA_ARCH__ >= 200
393 const int offset_minus_4 = 4 - (block_len & 3);
395 const int selector = (0x76543210 >> (offset_minus_4 * 4)) & 0xffff;
397 tmp0 = __byte_perm ( 0, append[0], selector);
398 tmp1 = __byte_perm (append[0], append[1], selector);
399 tmp2 = __byte_perm (append[1], 0, selector);
403 const u32 mod = block_len & 3;
407 case 0: tmp0 = append[0];
411 case 1: tmp0 = append[0] << 8;
412 tmp1 = append[0] >> 24 | append[1] << 8;
413 tmp2 = append[1] >> 24;
415 case 2: tmp0 = append[0] << 16;
416 tmp1 = append[0] >> 16 | append[1] << 16;
417 tmp2 = append[1] >> 16;
419 case 3: tmp0 = append[0] << 24;
420 tmp1 = append[0] >> 8 | append[1] << 24;
421 tmp2 = append[1] >> 8;
427 const u32 div = block_len / 4;
431 case 0: block0[0] |= tmp0;
435 case 1: block0[1] |= tmp0;
439 case 2: block0[2] |= tmp0;
443 case 3: block0[3] |= tmp0;
447 case 4: block1[0] |= tmp0;
451 case 5: block1[1] |= tmp0;
455 case 6: block1[2] |= tmp0;
459 case 7: block1[3] |= tmp0;
463 case 8: block2[0] |= tmp0;
467 case 9: block2[1] |= tmp0;
471 case 10: block2[2] |= tmp0;
475 case 11: block2[3] |= tmp0;
484 __device__ static void append_sign (u32x block0[4], u32x block1[4], const u32 block_len)
489 block0[0] = md5apr1_magic0;
490 block0[1] = md5apr1_magic1;
494 block0[0] = block0[0] | md5apr1_magic0 << 8;
495 block0[1] = md5apr1_magic0 >> 24 | md5apr1_magic1 << 8;
496 block0[2] = md5apr1_magic1 >> 24;
500 block0[0] = block0[0] | md5apr1_magic0 << 16;
501 block0[1] = md5apr1_magic0 >> 16 | md5apr1_magic1 << 16;
502 block0[2] = md5apr1_magic1 >> 16;
506 block0[0] = block0[0] | md5apr1_magic0 << 24;
507 block0[1] = md5apr1_magic0 >> 8 | md5apr1_magic1 << 24;
508 block0[2] = md5apr1_magic1 >> 8;
512 block0[1] = md5apr1_magic0;
513 block0[2] = md5apr1_magic1;
517 block0[1] = block0[1] | md5apr1_magic0 << 8;
518 block0[2] = md5apr1_magic0 >> 24 | md5apr1_magic1 << 8;
519 block0[3] = md5apr1_magic1 >> 24;
523 block0[1] = block0[1] | md5apr1_magic0 << 16;
524 block0[2] = md5apr1_magic0 >> 16 | md5apr1_magic1 << 16;
525 block0[3] = md5apr1_magic1 >> 16;
529 block0[1] = block0[1] | md5apr1_magic0 << 24;
530 block0[2] = md5apr1_magic0 >> 8 | md5apr1_magic1 << 24;
531 block0[3] = md5apr1_magic1 >> 8;
535 block0[2] = md5apr1_magic0;
536 block0[3] = md5apr1_magic1;
540 block0[2] = block0[2] | md5apr1_magic0 << 8;
541 block0[3] = md5apr1_magic0 >> 24 | md5apr1_magic1 << 8;
542 block1[0] = md5apr1_magic1 >> 24;
546 block0[2] = block0[2] | md5apr1_magic0 << 16;
547 block0[3] = md5apr1_magic0 >> 16 | md5apr1_magic1 << 16;
548 block1[0] = md5apr1_magic1 >> 16;
552 block0[2] = block0[2] | md5apr1_magic0 << 24;
553 block0[3] = md5apr1_magic0 >> 8 | md5apr1_magic1 << 24;
554 block1[0] = md5apr1_magic1 >> 8;
558 block0[3] = md5apr1_magic0;
559 block1[0] = md5apr1_magic1;
563 block0[3] = block0[3] | md5apr1_magic0 << 8;
564 block1[0] = md5apr1_magic0 >> 24 | md5apr1_magic1 << 8;
565 block1[1] = md5apr1_magic1 >> 24;
569 block0[3] = block0[3] | md5apr1_magic0 << 16;
570 block1[0] = md5apr1_magic0 >> 16 | md5apr1_magic1 << 16;
571 block1[1] = md5apr1_magic1 >> 16;
575 block0[3] = block0[3] | md5apr1_magic0 << 24;
576 block1[0] = md5apr1_magic0 >> 8 | md5apr1_magic1 << 24;
577 block1[1] = md5apr1_magic1 >> 8;
582 __device__ static void append_1st (u32x block0[4], u32x block1[4], u32x block2[4], u32x block3[4], const u32 block_len, const u32x append)
591 block0[0] = block0[0] | append << 8;
595 block0[0] = block0[0] | append << 16;
599 block0[0] = block0[0] | append << 24;
607 block0[1] = block0[1] | append << 8;
611 block0[1] = block0[1] | append << 16;
615 block0[1] = block0[1] | append << 24;
623 block0[2] = block0[2] | append << 8;
627 block0[2] = block0[2] | append << 16;
631 block0[2] = block0[2] | append << 24;
639 block0[3] = block0[3] | append << 8;
643 block0[3] = block0[3] | append << 16;
647 block0[3] = block0[3] | append << 24;
655 block1[0] = block1[0] | append << 8;
659 block1[0] = block1[0] | append << 16;
663 block1[0] = block1[0] | append << 24;
671 block1[1] = block1[1] | append << 8;
675 block1[1] = block1[1] | append << 16;
679 block1[1] = block1[1] | append << 24;
687 block1[2] = block1[2] | append << 8;
691 block1[2] = block1[2] | append << 16;
695 block1[2] = block1[2] | append << 24;
703 block1[3] = block1[3] | append << 8;
707 block1[3] = block1[3] | append << 16;
711 block1[3] = block1[3] | append << 24;
719 block2[0] = block2[0] | append << 8;
723 block2[0] = block2[0] | append << 16;
727 block2[0] = block2[0] | append << 24;
735 block2[1] = block2[1] | append << 8;
739 block2[1] = block2[1] | append << 16;
743 block2[1] = block2[1] | append << 24;
751 block2[2] = block2[2] | append << 8;
755 block2[2] = block2[2] | append << 16;
759 block2[2] = block2[2] | append << 24;
767 block2[3] = block2[3] | append << 8;
771 block2[3] = block2[3] | append << 16;
775 block2[3] = block2[3] | append << 24;
783 block3[0] = block3[0] | append << 8;
787 block3[0] = block3[0] | append << 16;
791 block3[0] = block3[0] | append << 24;
799 block3[1] = block3[1] | append << 8;
803 block3[1] = block3[1] | append << 16;
807 block3[1] = block3[1] | append << 24;
816 extern "C" __global__ void __launch_bounds__ (256, 1) m01600_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)
818 const u32 gid = (blockIdx.x * blockDim.x) + threadIdx.x;
820 if (gid >= gid_max) return;
824 w0[0] = pws[gid].i[0];
825 w0[1] = pws[gid].i[1];
826 w0[2] = pws[gid].i[2];
827 w0[3] = pws[gid].i[3];
829 const u32 pw_len = pws[gid].pw_len;
837 salt_buf[0] = salt_bufs[salt_pos].salt_buf[0];
838 salt_buf[1] = salt_bufs[salt_pos].salt_buf[1];
840 const u32 salt_len = salt_bufs[salt_pos].salt_len;
846 //memcat16 (block0, block1, block2, block3, block_len, w0);
847 //block_len += pw_len;
849 u32 block_len = pw_len;
879 memcat8 (block0, block1, block2, block3, block_len, salt_buf);
881 block_len += salt_len;
883 memcat16 (block0, block1, block2, block3, block_len, w0);
887 append_0x80_4 (block0, block1, block2, block3, block_len);
889 block3[2] = block_len * 8;
898 md5_transform (block0, block1, block2, block3, digest);
900 /* The password first, since that is what is most unknown */
901 /* Then our magic string */
902 /* Then the raw salt */
903 /* Then just as many characters of the MD5(pw,salt,pw) */
905 //memcat16 (block0, block1, block2, block3, block_len, w);
906 //block_len += pw_len;
930 append_sign (block0, block1, block_len);
934 memcat8 (block0, block1, block2, block3, block_len, salt_buf);
936 block_len += salt_len;
938 truncate_block (digest, pw_len);
940 memcat16 (block0, block1, block2, block3, block_len, digest);
944 /* Then something really weird... */
946 u32x append = block0[0] & 0xFF;
948 for (u32 j = pw_len; j; j >>= 1)
952 append_1st (block0, block1, block2, block3, block_len, append);
958 append_0x80_4 (block0, block1, block2, block3, block_len);
960 block3[2] = block_len * 8;
967 md5_transform (block0, block1, block2, block3, digest);
969 tmps[gid].digest_buf[0] = digest[0];
970 tmps[gid].digest_buf[1] = digest[1];
971 tmps[gid].digest_buf[2] = digest[2];
972 tmps[gid].digest_buf[3] = digest[3];
975 extern "C" __global__ void __launch_bounds__ (256, 1) m01600_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)
981 const u32 gid = (blockIdx.x * blockDim.x) + threadIdx.x;
983 if (gid >= gid_max) return;
987 w0[0] = pws[gid].i[0];
988 w0[1] = pws[gid].i[1];
989 w0[2] = pws[gid].i[2];
990 w0[3] = pws[gid].i[3];
992 const u32 pw_len = pws[gid].pw_len;
1001 append_0x80_1 (w0_x80, pw_len);
1009 salt_buf[0] = salt_bufs[salt_pos].salt_buf[0];
1010 salt_buf[1] = salt_bufs[salt_pos].salt_buf[1];
1012 const u32 salt_len = salt_bufs[salt_pos].salt_len;
1020 digest[0] = tmps[gid].digest_buf[0];
1021 digest[1] = tmps[gid].digest_buf[1];
1022 digest[2] = tmps[gid].digest_buf[2];
1023 digest[3] = tmps[gid].digest_buf[3];
1029 /* and now, just to make sure things don't run too fast */
1061 for (u32 i = 0, j = loop_pos; i < loop_cnt; i++, j++)
1074 const u32 j1 = (j & 1) ? 1 : 0;
1075 const u32 j3 = (j % 3) ? 1 : 0;
1076 const u32 j7 = (j % 7) ? 1 : 0;
1089 memcat8 (block0, block1, block2, block3, block_len, salt_buf);
1091 block_len += salt_len;
1096 memcat16 (block0, block1, block2, block3, block_len, w0);
1098 block_len += pw_len;
1101 memcat16_x80 (block0, block1, block2, block3, block_len, digest);
1107 block0[0] = digest[0];
1108 block0[1] = digest[1];
1109 block0[2] = digest[2];
1110 block0[3] = digest[3];
1116 block1[0] = salt_buf[0];
1117 block1[1] = salt_buf[1];
1119 block_len += salt_len;
1121 memcat16 (block0, block1, block2, block3, block_len, w0);
1123 block_len += pw_len;
1127 block1[0] = salt_buf[0];
1128 block1[1] = salt_buf[1];
1130 block_len += salt_len;
1139 block_len += pw_len;
1142 memcat16 (block0, block1, block2, block3, block_len, w0_x80);
1144 block_len += pw_len;
1147 block3[2] = block_len * 8;
1154 md5_transform (block0, block1, block2, block3, digest);
1157 tmps[gid].digest_buf[0] = digest[0];
1158 tmps[gid].digest_buf[1] = digest[1];
1159 tmps[gid].digest_buf[2] = digest[2];
1160 tmps[gid].digest_buf[3] = digest[3];
1163 extern "C" __global__ void __launch_bounds__ (256, 1) m01600_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)
1169 const u32 gid = (blockIdx.x * blockDim.x) + threadIdx.x;
1171 if (gid >= gid_max) return;
1173 const u32 lid = threadIdx.x;
1179 const u32x r0 = tmps[gid].digest_buf[DGST_R0];
1180 const u32x r1 = tmps[gid].digest_buf[DGST_R1];
1181 const u32x r2 = tmps[gid].digest_buf[DGST_R2];
1182 const u32x r3 = tmps[gid].digest_buf[DGST_R3];
1186 #include VECT_COMPARE_M