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"
20 #define COMPARE_S "OpenCL/check_single_comp4.c"
21 #define COMPARE_M "OpenCL/check_multi_comp4.c"
23 __constant u32 k_sha256[64] =
25 SHA256C00, SHA256C01, SHA256C02, SHA256C03,
26 SHA256C04, SHA256C05, SHA256C06, SHA256C07,
27 SHA256C08, SHA256C09, SHA256C0a, SHA256C0b,
28 SHA256C0c, SHA256C0d, SHA256C0e, SHA256C0f,
29 SHA256C10, SHA256C11, SHA256C12, SHA256C13,
30 SHA256C14, SHA256C15, SHA256C16, SHA256C17,
31 SHA256C18, SHA256C19, SHA256C1a, SHA256C1b,
32 SHA256C1c, SHA256C1d, SHA256C1e, SHA256C1f,
33 SHA256C20, SHA256C21, SHA256C22, SHA256C23,
34 SHA256C24, SHA256C25, SHA256C26, SHA256C27,
35 SHA256C28, SHA256C29, SHA256C2a, SHA256C2b,
36 SHA256C2c, SHA256C2d, SHA256C2e, SHA256C2f,
37 SHA256C30, SHA256C31, SHA256C32, SHA256C33,
38 SHA256C34, SHA256C35, SHA256C36, SHA256C37,
39 SHA256C38, SHA256C39, SHA256C3a, SHA256C3b,
40 SHA256C3c, SHA256C3d, SHA256C3e, SHA256C3f,
43 static void sha256_transform (const u32 w[16], u32 digest[8])
54 u32 w0_t = swap32 (w[ 0]);
55 u32 w1_t = swap32 (w[ 1]);
56 u32 w2_t = swap32 (w[ 2]);
57 u32 w3_t = swap32 (w[ 3]);
58 u32 w4_t = swap32 (w[ 4]);
59 u32 w5_t = swap32 (w[ 5]);
60 u32 w6_t = swap32 (w[ 6]);
61 u32 w7_t = swap32 (w[ 7]);
62 u32 w8_t = swap32 (w[ 8]);
63 u32 w9_t = swap32 (w[ 9]);
64 u32 wa_t = swap32 (w[10]);
65 u32 wb_t = swap32 (w[11]);
66 u32 wc_t = swap32 (w[12]);
67 u32 wd_t = swap32 (w[13]);
68 u32 we_t = swap32 (w[14]);
69 u32 wf_t = swap32 (w[15]);
71 #define ROUND_EXPAND() \
73 w0_t = SHA256_EXPAND (we_t, w9_t, w1_t, w0_t); \
74 w1_t = SHA256_EXPAND (wf_t, wa_t, w2_t, w1_t); \
75 w2_t = SHA256_EXPAND (w0_t, wb_t, w3_t, w2_t); \
76 w3_t = SHA256_EXPAND (w1_t, wc_t, w4_t, w3_t); \
77 w4_t = SHA256_EXPAND (w2_t, wd_t, w5_t, w4_t); \
78 w5_t = SHA256_EXPAND (w3_t, we_t, w6_t, w5_t); \
79 w6_t = SHA256_EXPAND (w4_t, wf_t, w7_t, w6_t); \
80 w7_t = SHA256_EXPAND (w5_t, w0_t, w8_t, w7_t); \
81 w8_t = SHA256_EXPAND (w6_t, w1_t, w9_t, w8_t); \
82 w9_t = SHA256_EXPAND (w7_t, w2_t, wa_t, w9_t); \
83 wa_t = SHA256_EXPAND (w8_t, w3_t, wb_t, wa_t); \
84 wb_t = SHA256_EXPAND (w9_t, w4_t, wc_t, wb_t); \
85 wc_t = SHA256_EXPAND (wa_t, w5_t, wd_t, wc_t); \
86 wd_t = SHA256_EXPAND (wb_t, w6_t, we_t, wd_t); \
87 we_t = SHA256_EXPAND (wc_t, w7_t, wf_t, we_t); \
88 wf_t = SHA256_EXPAND (wd_t, w8_t, w0_t, wf_t); \
91 #define ROUND_STEP(i) \
93 SHA256_STEP (SHA256_F0o, SHA256_F1o, a, b, c, d, e, f, g, h, w0_t, k_sha256[i + 0]); \
94 SHA256_STEP (SHA256_F0o, SHA256_F1o, h, a, b, c, d, e, f, g, w1_t, k_sha256[i + 1]); \
95 SHA256_STEP (SHA256_F0o, SHA256_F1o, g, h, a, b, c, d, e, f, w2_t, k_sha256[i + 2]); \
96 SHA256_STEP (SHA256_F0o, SHA256_F1o, f, g, h, a, b, c, d, e, w3_t, k_sha256[i + 3]); \
97 SHA256_STEP (SHA256_F0o, SHA256_F1o, e, f, g, h, a, b, c, d, w4_t, k_sha256[i + 4]); \
98 SHA256_STEP (SHA256_F0o, SHA256_F1o, d, e, f, g, h, a, b, c, w5_t, k_sha256[i + 5]); \
99 SHA256_STEP (SHA256_F0o, SHA256_F1o, c, d, e, f, g, h, a, b, w6_t, k_sha256[i + 6]); \
100 SHA256_STEP (SHA256_F0o, SHA256_F1o, b, c, d, e, f, g, h, a, w7_t, k_sha256[i + 7]); \
101 SHA256_STEP (SHA256_F0o, SHA256_F1o, a, b, c, d, e, f, g, h, w8_t, k_sha256[i + 8]); \
102 SHA256_STEP (SHA256_F0o, SHA256_F1o, h, a, b, c, d, e, f, g, w9_t, k_sha256[i + 9]); \
103 SHA256_STEP (SHA256_F0o, SHA256_F1o, g, h, a, b, c, d, e, f, wa_t, k_sha256[i + 10]); \
104 SHA256_STEP (SHA256_F0o, SHA256_F1o, f, g, h, a, b, c, d, e, wb_t, k_sha256[i + 11]); \
105 SHA256_STEP (SHA256_F0o, SHA256_F1o, e, f, g, h, a, b, c, d, wc_t, k_sha256[i + 12]); \
106 SHA256_STEP (SHA256_F0o, SHA256_F1o, d, e, f, g, h, a, b, c, wd_t, k_sha256[i + 13]); \
107 SHA256_STEP (SHA256_F0o, SHA256_F1o, c, d, e, f, g, h, a, b, we_t, k_sha256[i + 14]); \
108 SHA256_STEP (SHA256_F0o, SHA256_F1o, b, c, d, e, f, g, h, a, wf_t, k_sha256[i + 15]); \
114 for (int i = 16; i < 64; i += 16)
116 ROUND_EXPAND (); ROUND_STEP (i);
129 static void sha256_transform_no14 (const u32 w[16], u32 digest[8])
150 sha256_transform (w_t, digest);
153 static void init_ctx (u32 digest[8])
155 digest[0] = SHA256M_A;
156 digest[1] = SHA256M_B;
157 digest[2] = SHA256M_C;
158 digest[3] = SHA256M_D;
159 digest[4] = SHA256M_E;
160 digest[5] = SHA256M_F;
161 digest[6] = SHA256M_G;
162 digest[7] = SHA256M_H;
165 static void bzero16 (u32 block[16])
185 static void bswap8 (u32 block[16])
187 block[ 0] = swap32 (block[ 0]);
188 block[ 1] = swap32 (block[ 1]);
189 block[ 2] = swap32 (block[ 2]);
190 block[ 3] = swap32 (block[ 3]);
191 block[ 4] = swap32 (block[ 4]);
192 block[ 5] = swap32 (block[ 5]);
193 block[ 6] = swap32 (block[ 6]);
194 block[ 7] = swap32 (block[ 7]);
197 static u32 memcat16 (u32 block[16], const u32 block_len, const u32 append[4], const u32 append_len)
199 const u32 mod = block_len & 3;
200 const u32 div = block_len / 4;
208 #if defined IS_AMD || defined IS_GENERIC
209 const int offset_minus_4 = 4 - block_len;
211 tmp0 = amd_bytealign (append[0], 0, offset_minus_4);
212 tmp1 = amd_bytealign (append[1], append[0], offset_minus_4);
213 tmp2 = amd_bytealign (append[2], append[1], offset_minus_4);
214 tmp3 = amd_bytealign (append[3], append[2], offset_minus_4);
215 tmp4 = amd_bytealign ( 0, append[3], offset_minus_4);
228 const int offset_minus_4 = 4 - (block_len & 3);
230 const int selector = (0x76543210 >> (offset_minus_4 * 4)) & 0xffff;
232 tmp0 = __byte_perm ( 0, append[0], selector);
233 tmp1 = __byte_perm (append[0], append[1], selector);
234 tmp2 = __byte_perm (append[1], append[2], selector);
235 tmp3 = __byte_perm (append[2], append[3], selector);
236 tmp4 = __byte_perm (append[3], 0, selector);
241 case 0: block[ 0] |= tmp0;
247 case 1: block[ 1] |= tmp0;
253 case 2: block[ 2] |= tmp0;
259 case 3: block[ 3] |= tmp0;
265 case 4: block[ 4] |= tmp0;
271 case 5: block[ 5] |= tmp0;
277 case 6: block[ 6] |= tmp0;
283 case 7: block[ 7] |= tmp0;
289 case 8: block[ 8] |= tmp0;
295 case 9: block[ 9] |= tmp0;
301 case 10: block[10] |= tmp0;
307 case 11: block[11] |= tmp0;
313 case 12: block[12] |= tmp0;
318 case 13: block[13] |= tmp0;
322 case 14: block[14] |= tmp0;
325 case 15: block[15] |= tmp0;
329 u32 new_len = block_len + append_len;
334 static u32 memcat16c (u32 block[16], const u32 block_len, const u32 append[4], const u32 append_len, u32 digest[8])
336 const u32 mod = block_len & 3;
337 const u32 div = block_len / 4;
345 #if defined IS_AMD || defined IS_GENERIC
346 const int offset_minus_4 = 4 - block_len;
348 tmp0 = amd_bytealign (append[0], 0, offset_minus_4);
349 tmp1 = amd_bytealign (append[1], append[0], offset_minus_4);
350 tmp2 = amd_bytealign (append[2], append[1], offset_minus_4);
351 tmp3 = amd_bytealign (append[3], append[2], offset_minus_4);
352 tmp4 = amd_bytealign ( 0, append[3], offset_minus_4);
365 const int offset_minus_4 = 4 - (block_len & 3);
367 const int selector = (0x76543210 >> (offset_minus_4 * 4)) & 0xffff;
369 tmp0 = __byte_perm ( 0, append[0], selector);
370 tmp1 = __byte_perm (append[0], append[1], selector);
371 tmp2 = __byte_perm (append[1], append[2], selector);
372 tmp3 = __byte_perm (append[2], append[3], selector);
373 tmp4 = __byte_perm (append[3], 0, selector);
376 u32 carry[4] = { 0, 0, 0, 0 };
380 case 0: block[ 0] |= tmp0;
386 case 1: block[ 1] |= tmp0;
392 case 2: block[ 2] |= tmp0;
398 case 3: block[ 3] |= tmp0;
404 case 4: block[ 4] |= tmp0;
410 case 5: block[ 5] |= tmp0;
416 case 6: block[ 6] |= tmp0;
422 case 7: block[ 7] |= tmp0;
428 case 8: block[ 8] |= tmp0;
434 case 9: block[ 9] |= tmp0;
440 case 10: block[10] |= tmp0;
446 case 11: block[11] |= tmp0;
452 case 12: block[12] |= tmp0;
458 case 13: block[13] |= tmp0;
464 case 14: block[14] |= tmp0;
470 case 15: block[15] |= tmp0;
478 u32 new_len = block_len + append_len;
484 sha256_transform (block, digest);
497 static u32 memcat20 (u32 block[20], const u32 block_len, const u32 append[4], const u32 append_len)
499 const u32 mod = block_len & 3;
500 const u32 div = block_len / 4;
508 #if defined IS_AMD || defined IS_GENERIC
509 const int offset_minus_4 = 4 - block_len;
511 tmp0 = amd_bytealign (append[0], 0, offset_minus_4);
512 tmp1 = amd_bytealign (append[1], append[0], offset_minus_4);
513 tmp2 = amd_bytealign (append[2], append[1], offset_minus_4);
514 tmp3 = amd_bytealign (append[3], append[2], offset_minus_4);
515 tmp4 = amd_bytealign ( 0, append[3], offset_minus_4);
528 const int offset_minus_4 = 4 - (block_len & 3);
530 const int selector = (0x76543210 >> (offset_minus_4 * 4)) & 0xffff;
532 tmp0 = __byte_perm ( 0, append[0], selector);
533 tmp1 = __byte_perm (append[0], append[1], selector);
534 tmp2 = __byte_perm (append[1], append[2], selector);
535 tmp3 = __byte_perm (append[2], append[3], selector);
536 tmp4 = __byte_perm (append[3], 0, selector);
541 case 0: block[ 0] |= tmp0;
547 case 1: block[ 1] |= tmp0;
553 case 2: block[ 2] |= tmp0;
559 case 3: block[ 3] |= tmp0;
565 case 4: block[ 4] |= tmp0;
571 case 5: block[ 5] |= tmp0;
577 case 6: block[ 6] |= tmp0;
583 case 7: block[ 7] |= tmp0;
589 case 8: block[ 8] |= tmp0;
595 case 9: block[ 9] |= tmp0;
601 case 10: block[10] |= tmp0;
607 case 11: block[11] |= tmp0;
613 case 12: block[12] |= tmp0;
619 case 13: block[13] |= tmp0;
625 case 14: block[14] |= tmp0;
631 case 15: block[15] |= tmp0;
639 return block_len + append_len;
642 static u32 memcat20_x80 (u32 block[20], const u32 block_len, const u32 append[4], const u32 append_len)
644 const u32 mod = block_len & 3;
645 const u32 div = block_len / 4;
653 #if defined IS_AMD || defined IS_GENERIC
654 const int offset_minus_4 = 4 - block_len;
656 tmp0 = amd_bytealign (append[0], 0, offset_minus_4);
657 tmp1 = amd_bytealign (append[1], append[0], offset_minus_4);
658 tmp2 = amd_bytealign (append[2], append[1], offset_minus_4);
659 tmp3 = amd_bytealign (append[3], append[2], offset_minus_4);
660 tmp4 = amd_bytealign ( 0x80, append[3], offset_minus_4);
673 const int offset_minus_4 = 4 - (block_len & 3);
675 const int selector = (0x76543210 >> (offset_minus_4 * 4)) & 0xffff;
677 tmp0 = __byte_perm ( 0, append[0], selector);
678 tmp1 = __byte_perm (append[0], append[1], selector);
679 tmp2 = __byte_perm (append[1], append[2], selector);
680 tmp3 = __byte_perm (append[2], append[3], selector);
681 tmp4 = __byte_perm (append[3], 0x80, selector);
686 case 0: block[ 0] |= tmp0;
692 case 1: block[ 1] |= tmp0;
698 case 2: block[ 2] |= tmp0;
704 case 3: block[ 3] |= tmp0;
710 case 4: block[ 4] |= tmp0;
716 case 5: block[ 5] |= tmp0;
722 case 6: block[ 6] |= tmp0;
728 case 7: block[ 7] |= tmp0;
734 case 8: block[ 8] |= tmp0;
740 case 9: block[ 9] |= tmp0;
746 case 10: block[10] |= tmp0;
752 case 11: block[11] |= tmp0;
758 case 12: block[12] |= tmp0;
764 case 13: block[13] |= tmp0;
770 case 14: block[14] |= tmp0;
776 case 15: block[15] |= tmp0;
784 return block_len + append_len;
787 __kernel void m07400_init (__global pw_t *pws, __global kernel_rule_t *rules_buf, __global comb_t *combs_buf, __global bf_t *bfs_buf, __global sha256crypt_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 il_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
793 const u32 gid = get_global_id (0);
795 if (gid >= gid_max) return;
799 w0[0] = pws[gid].i[0];
800 w0[1] = pws[gid].i[1];
801 w0[2] = pws[gid].i[2];
802 w0[3] = pws[gid].i[3];
804 const u32 pw_len = pws[gid].pw_len;
812 salt_buf[0] = salt_bufs[salt_pos].salt_buf[0];
813 salt_buf[1] = salt_bufs[salt_pos].salt_buf[1];
814 salt_buf[2] = salt_bufs[salt_pos].salt_buf[2];
815 salt_buf[3] = salt_bufs[salt_pos].salt_buf[3];
817 u32 salt_len = salt_bufs[salt_pos].salt_len;
823 u32 block_len; // never reaches > 64
824 u32 transform_len; // required for w[15] = len * 8
832 /* Prepare for the real work. */
840 block_len = memcat16 (block, block_len, w0, pw_len);
844 block_len = memcat16 (block, block_len, salt_buf, salt_len);
848 block_len = memcat16 (block, block_len, w0, pw_len);
850 append_0x80_1x16 (block, block_len);
852 block[15] = swap32 (block_len * 8);
854 init_ctx (alt_result);
856 sha256_transform (block, alt_result);
864 u32 alt_result_tmp[8];
866 alt_result_tmp[0] = alt_result[0];
867 alt_result_tmp[1] = alt_result[1];
868 alt_result_tmp[2] = alt_result[2];
869 alt_result_tmp[3] = alt_result[3];
870 alt_result_tmp[4] = 0;
871 alt_result_tmp[5] = 0;
872 alt_result_tmp[6] = 0;
873 alt_result_tmp[7] = 0;
875 truncate_block (alt_result_tmp, pw_len);
877 /* Add the key string. */
879 block_len = memcat16 (block, block_len, w0, pw_len);
881 /* The last part is the salt string. This must be at most 8
882 characters and it ends at the first `$' character (for
883 compatibility with existing implementations). */
885 block_len = memcat16 (block, block_len, salt_buf, salt_len);
887 /* Now get result of this (32 bytes) and add it to the other
890 block_len = memcat16 (block, block_len, alt_result_tmp, pw_len);
892 transform_len = block_len;
894 /* Take the binary representation of the length of the key and for every
895 1 add the alternate sum, for every 0 the key. */
897 alt_result_tmp[0] = alt_result[0];
898 alt_result_tmp[1] = alt_result[1];
899 alt_result_tmp[2] = alt_result[2];
900 alt_result_tmp[3] = alt_result[3];
901 alt_result_tmp[4] = alt_result[4];
902 alt_result_tmp[5] = alt_result[5];
903 alt_result_tmp[6] = alt_result[6];
904 alt_result_tmp[7] = alt_result[7];
906 init_ctx (alt_result);
908 for (u32 j = pw_len; j; j >>= 1)
912 block_len = memcat16c (block, block_len, &alt_result_tmp[0], 16, alt_result);
913 block_len = memcat16c (block, block_len, &alt_result_tmp[4], 16, alt_result);
919 block_len = memcat16c (block, block_len, w0, pw_len, alt_result);
921 transform_len += pw_len;
925 append_0x80_1x16 (block, block_len);
929 sha256_transform (block, alt_result);
934 block[15] = swap32 (transform_len * 8);
936 sha256_transform (block, alt_result);
940 tmps[gid].alt_result[0] = alt_result[0];
941 tmps[gid].alt_result[1] = alt_result[1];
942 tmps[gid].alt_result[2] = alt_result[2];
943 tmps[gid].alt_result[3] = alt_result[3];
944 tmps[gid].alt_result[4] = alt_result[4];
945 tmps[gid].alt_result[5] = alt_result[5];
946 tmps[gid].alt_result[6] = alt_result[6];
947 tmps[gid].alt_result[7] = alt_result[7];
949 /* Start computation of P byte sequence. */
957 /* For every character in the password add the entire password. */
961 for (u32 j = 0; j < pw_len; j++)
963 block_len = memcat16c (block, block_len, w0, pw_len, p_bytes);
965 transform_len += pw_len;
968 /* Finish the digest. */
970 append_0x80_1x16 (block, block_len);
974 sha256_transform (block, p_bytes);
979 block[15] = swap32 (transform_len * 8);
981 sha256_transform (block, p_bytes);
985 truncate_block (p_bytes, pw_len);
987 tmps[gid].p_bytes[0] = p_bytes[0];
988 tmps[gid].p_bytes[1] = p_bytes[1];
989 tmps[gid].p_bytes[2] = p_bytes[2];
990 tmps[gid].p_bytes[3] = p_bytes[3];
992 /* Start computation of S byte sequence. */
1000 /* For every character in the password add the entire password. */
1004 for (u32 j = 0; j < 16 + (alt_result[0] & 0xff); j++)
1006 block_len = memcat16c (block, block_len, salt_buf, salt_len, s_bytes);
1008 transform_len += salt_len;
1011 /* Finish the digest. */
1013 append_0x80_1x16 (block, block_len);
1015 if (block_len >= 56)
1017 sha256_transform (block, s_bytes);
1022 block[15] = swap32 (transform_len * 8);
1024 sha256_transform (block, s_bytes);
1028 truncate_block (s_bytes, salt_len);
1030 tmps[gid].s_bytes[0] = s_bytes[0];
1031 tmps[gid].s_bytes[1] = s_bytes[1];
1032 tmps[gid].s_bytes[2] = s_bytes[2];
1033 tmps[gid].s_bytes[3] = s_bytes[3];
1036 __kernel void m07400_loop (__global pw_t *pws, __global kernel_rule_t *rules_buf, __global comb_t *combs_buf, __global bf_t *bfs_buf, __global sha256crypt_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 il_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
1042 const u32 gid = get_global_id (0);
1044 if (gid >= gid_max) return;
1046 const u32 pw_len = pws[gid].pw_len;
1054 p_bytes[0] = tmps[gid].p_bytes[0];
1055 p_bytes[1] = tmps[gid].p_bytes[1];
1056 p_bytes[2] = tmps[gid].p_bytes[2];
1057 p_bytes[3] = tmps[gid].p_bytes[3];
1061 p_bytes_x80[0] = tmps[gid].p_bytes[0];
1062 p_bytes_x80[1] = tmps[gid].p_bytes[1];
1063 p_bytes_x80[2] = tmps[gid].p_bytes[2];
1064 p_bytes_x80[3] = tmps[gid].p_bytes[3];
1066 append_0x80_1x4 (p_bytes_x80, pw_len);
1070 s_bytes[0] = tmps[gid].s_bytes[0];
1071 s_bytes[1] = tmps[gid].s_bytes[1];
1072 s_bytes[2] = tmps[gid].s_bytes[2];
1073 s_bytes[3] = tmps[gid].s_bytes[3];
1077 alt_result[0] = tmps[gid].alt_result[0];
1078 alt_result[1] = tmps[gid].alt_result[1];
1079 alt_result[2] = tmps[gid].alt_result[2];
1080 alt_result[3] = tmps[gid].alt_result[3];
1081 alt_result[4] = tmps[gid].alt_result[4];
1082 alt_result[5] = tmps[gid].alt_result[5];
1083 alt_result[6] = tmps[gid].alt_result[6];
1084 alt_result[7] = tmps[gid].alt_result[7];
1086 u32 salt_len = salt_bufs[salt_pos].salt_len;
1088 /* Repeatedly run the collected hash value through SHA256 to burn
1091 for (u32 i = 0, j = loop_pos; i < loop_cnt; i++, j++)
1099 bzero16 (&block[ 0]);
1100 bzero16 (&block[16]);
1104 const u32 j1 = (j & 1) ? 1 : 0;
1105 const u32 j3 = (j % 3) ? 1 : 0;
1106 const u32 j7 = (j % 7) ? 1 : 0;
1110 block[0] = p_bytes[0];
1111 block[1] = p_bytes[1];
1112 block[2] = p_bytes[2];
1113 block[3] = p_bytes[3];
1119 block[0] = alt_result[0];
1120 block[1] = alt_result[1];
1121 block[2] = alt_result[2];
1122 block[3] = alt_result[3];
1123 block[4] = alt_result[4];
1124 block[5] = alt_result[5];
1125 block[6] = alt_result[6];
1126 block[7] = alt_result[7];
1133 block_len = memcat20 (block, block_len, s_bytes, salt_len);
1138 block_len = memcat20 (block, block_len, p_bytes, pw_len);
1143 block_len = memcat20 (block, block_len, &alt_result[0], 16);
1144 block_len = memcat20_x80 (block, block_len, &alt_result[4], 16);
1148 block_len = memcat20 (block, block_len, p_bytes_x80, pw_len);
1151 if (block_len >= 56)
1153 sha256_transform (block, tmp);
1155 block[ 0] = block[16];
1156 block[ 1] = block[17];
1157 block[ 2] = block[18];
1158 block[ 3] = block[19];
1173 block[15] = swap32 (block_len * 8);
1175 sha256_transform_no14 (block, tmp);
1179 alt_result[0] = tmp[0];
1180 alt_result[1] = tmp[1];
1181 alt_result[2] = tmp[2];
1182 alt_result[3] = tmp[3];
1183 alt_result[4] = tmp[4];
1184 alt_result[5] = tmp[5];
1185 alt_result[6] = tmp[6];
1186 alt_result[7] = tmp[7];
1189 tmps[gid].alt_result[0] = alt_result[0];
1190 tmps[gid].alt_result[1] = alt_result[1];
1191 tmps[gid].alt_result[2] = alt_result[2];
1192 tmps[gid].alt_result[3] = alt_result[3];
1193 tmps[gid].alt_result[4] = alt_result[4];
1194 tmps[gid].alt_result[5] = alt_result[5];
1195 tmps[gid].alt_result[6] = alt_result[6];
1196 tmps[gid].alt_result[7] = alt_result[7];
1199 __kernel void m07400_comp (__global pw_t *pws, __global kernel_rule_t *rules_buf, __global comb_t *combs_buf, __global bf_t *bfs_buf, __global sha256crypt_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 il_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
1205 const u32 gid = get_global_id (0);
1207 if (gid >= gid_max) return;
1209 const u32 lid = get_local_id (0);
1211 const u32 r0 = tmps[gid].alt_result[0];
1212 const u32 r1 = tmps[gid].alt_result[1];
1213 const u32 r2 = tmps[gid].alt_result[2];
1214 const u32 r3 = tmps[gid].alt_result[3];