/** * Author......: Jens Steube * License.....: MIT */ #define _DCC2_ #include "include/constants.h" #include "include/kernel_vendor.h" #ifdef VLIW1 #define VECT_SIZE1 #endif #ifdef VLIW4 #define VECT_SIZE2 #endif #ifdef VLIW5 #define VECT_SIZE2 #endif #define DGST_R0 0 #define DGST_R1 1 #define DGST_R2 2 #define DGST_R3 3 #include "include/kernel_functions.c" #include "types_amd.c" #include "common_amd.c" #ifdef VECT_SIZE1 #define VECT_COMPARE_M "check_multi_vect1_comp4.c" #endif #ifdef VECT_SIZE2 #define VECT_COMPARE_M "check_multi_vect2_comp4.c" #endif static void md4_transform (const u32x w0[4], const u32x w1[4], const u32x w2[4], const u32x w3[4], u32x digest[4]) { u32x a = digest[0]; u32x b = digest[1]; u32x c = digest[2]; u32x d = digest[3]; MD4_STEP (MD4_Fo, a, b, c, d, w0[0], MD4C00, MD4S00); MD4_STEP (MD4_Fo, d, a, b, c, w0[1], MD4C00, MD4S01); MD4_STEP (MD4_Fo, c, d, a, b, w0[2], MD4C00, MD4S02); MD4_STEP (MD4_Fo, b, c, d, a, w0[3], MD4C00, MD4S03); MD4_STEP (MD4_Fo, a, b, c, d, w1[0], MD4C00, MD4S00); MD4_STEP (MD4_Fo, d, a, b, c, w1[1], MD4C00, MD4S01); MD4_STEP (MD4_Fo, c, d, a, b, w1[2], MD4C00, MD4S02); MD4_STEP (MD4_Fo, b, c, d, a, w1[3], MD4C00, MD4S03); MD4_STEP (MD4_Fo, a, b, c, d, w2[0], MD4C00, MD4S00); MD4_STEP (MD4_Fo, d, a, b, c, w2[1], MD4C00, MD4S01); MD4_STEP (MD4_Fo, c, d, a, b, w2[2], MD4C00, MD4S02); MD4_STEP (MD4_Fo, b, c, d, a, w2[3], MD4C00, MD4S03); MD4_STEP (MD4_Fo, a, b, c, d, w3[0], MD4C00, MD4S00); MD4_STEP (MD4_Fo, d, a, b, c, w3[1], MD4C00, MD4S01); MD4_STEP (MD4_Fo, c, d, a, b, w3[2], MD4C00, MD4S02); MD4_STEP (MD4_Fo, b, c, d, a, w3[3], MD4C00, MD4S03); MD4_STEP (MD4_Go, a, b, c, d, w0[0], MD4C01, MD4S10); MD4_STEP (MD4_Go, d, a, b, c, w1[0], MD4C01, MD4S11); MD4_STEP (MD4_Go, c, d, a, b, w2[0], MD4C01, MD4S12); MD4_STEP (MD4_Go, b, c, d, a, w3[0], MD4C01, MD4S13); MD4_STEP (MD4_Go, a, b, c, d, w0[1], MD4C01, MD4S10); MD4_STEP (MD4_Go, d, a, b, c, w1[1], MD4C01, MD4S11); MD4_STEP (MD4_Go, c, d, a, b, w2[1], MD4C01, MD4S12); MD4_STEP (MD4_Go, b, c, d, a, w3[1], MD4C01, MD4S13); MD4_STEP (MD4_Go, a, b, c, d, w0[2], MD4C01, MD4S10); MD4_STEP (MD4_Go, d, a, b, c, w1[2], MD4C01, MD4S11); MD4_STEP (MD4_Go, c, d, a, b, w2[2], MD4C01, MD4S12); MD4_STEP (MD4_Go, b, c, d, a, w3[2], MD4C01, MD4S13); MD4_STEP (MD4_Go, a, b, c, d, w0[3], MD4C01, MD4S10); MD4_STEP (MD4_Go, d, a, b, c, w1[3], MD4C01, MD4S11); MD4_STEP (MD4_Go, c, d, a, b, w2[3], MD4C01, MD4S12); MD4_STEP (MD4_Go, b, c, d, a, w3[3], MD4C01, MD4S13); MD4_STEP (MD4_H , a, b, c, d, w0[0], MD4C02, MD4S20); MD4_STEP (MD4_H , d, a, b, c, w2[0], MD4C02, MD4S21); MD4_STEP (MD4_H , c, d, a, b, w1[0], MD4C02, MD4S22); MD4_STEP (MD4_H , b, c, d, a, w3[0], MD4C02, MD4S23); MD4_STEP (MD4_H , a, b, c, d, w0[2], MD4C02, MD4S20); MD4_STEP (MD4_H , d, a, b, c, w2[2], MD4C02, MD4S21); MD4_STEP (MD4_H , c, d, a, b, w1[2], MD4C02, MD4S22); MD4_STEP (MD4_H , b, c, d, a, w3[2], MD4C02, MD4S23); MD4_STEP (MD4_H , a, b, c, d, w0[1], MD4C02, MD4S20); MD4_STEP (MD4_H , d, a, b, c, w2[1], MD4C02, MD4S21); MD4_STEP (MD4_H , c, d, a, b, w1[1], MD4C02, MD4S22); MD4_STEP (MD4_H , b, c, d, a, w3[1], MD4C02, MD4S23); MD4_STEP (MD4_H , a, b, c, d, w0[3], MD4C02, MD4S20); MD4_STEP (MD4_H , d, a, b, c, w2[3], MD4C02, MD4S21); MD4_STEP (MD4_H , c, d, a, b, w1[3], MD4C02, MD4S22); MD4_STEP (MD4_H , b, c, d, a, w3[3], MD4C02, MD4S23); digest[0] += a; digest[1] += b; digest[2] += c; digest[3] += d; } static void sha1_transform (const u32x w0[4], const u32x w1[4], const u32x w2[4], const u32x w3[4], u32x digest[5]) { u32x A = digest[0]; u32x B = digest[1]; u32x C = digest[2]; u32x D = digest[3]; u32x E = digest[4]; u32x w0_t = w0[0]; u32x w1_t = w0[1]; u32x w2_t = w0[2]; u32x w3_t = w0[3]; u32x w4_t = w1[0]; u32x w5_t = w1[1]; u32x w6_t = w1[2]; u32x w7_t = w1[3]; u32x w8_t = w2[0]; u32x w9_t = w2[1]; u32x wa_t = w2[2]; u32x wb_t = w2[3]; u32x wc_t = w3[0]; u32x wd_t = w3[1]; u32x we_t = w3[2]; u32x wf_t = w3[3]; #undef K #define K SHA1C00 SHA1_STEP (SHA1_F0o, A, B, C, D, E, w0_t); SHA1_STEP (SHA1_F0o, E, A, B, C, D, w1_t); SHA1_STEP (SHA1_F0o, D, E, A, B, C, w2_t); SHA1_STEP (SHA1_F0o, C, D, E, A, B, w3_t); SHA1_STEP (SHA1_F0o, B, C, D, E, A, w4_t); SHA1_STEP (SHA1_F0o, A, B, C, D, E, w5_t); SHA1_STEP (SHA1_F0o, E, A, B, C, D, w6_t); SHA1_STEP (SHA1_F0o, D, E, A, B, C, w7_t); SHA1_STEP (SHA1_F0o, C, D, E, A, B, w8_t); SHA1_STEP (SHA1_F0o, B, C, D, E, A, w9_t); SHA1_STEP (SHA1_F0o, A, B, C, D, E, wa_t); SHA1_STEP (SHA1_F0o, E, A, B, C, D, wb_t); SHA1_STEP (SHA1_F0o, D, E, A, B, C, wc_t); SHA1_STEP (SHA1_F0o, C, D, E, A, B, wd_t); SHA1_STEP (SHA1_F0o, B, C, D, E, A, we_t); SHA1_STEP (SHA1_F0o, A, B, C, D, E, wf_t); w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F0o, E, A, B, C, D, w0_t); w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F0o, D, E, A, B, C, w1_t); w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F0o, C, D, E, A, B, w2_t); w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F0o, B, C, D, E, A, w3_t); #undef K #define K SHA1C01 w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w4_t); w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w5_t); w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w6_t); w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w7_t); w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w8_t); w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w9_t); wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wa_t); wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, wb_t); wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, wc_t); wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, wd_t); we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, we_t); wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wf_t); w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w0_t); w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w1_t); w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w2_t); w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w3_t); w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w4_t); w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w5_t); w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w6_t); w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w7_t); #undef K #define K SHA1C02 w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, w8_t); w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, w9_t); wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, wa_t); wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, wb_t); wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, wc_t); wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, wd_t); we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, we_t); wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, wf_t); w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, w0_t); w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, w1_t); w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, w2_t); w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, w3_t); w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, w4_t); w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, w5_t); w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, w6_t); w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F2o, A, B, C, D, E, w7_t); w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F2o, E, A, B, C, D, w8_t); w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F2o, D, E, A, B, C, w9_t); wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F2o, C, D, E, A, B, wa_t); wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F2o, B, C, D, E, A, wb_t); #undef K #define K SHA1C03 wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, wc_t); wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wd_t); we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, we_t); wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, wf_t); w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w0_t); w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w1_t); w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w2_t); w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w3_t); w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w4_t); w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, w5_t); w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, w6_t); w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, w7_t); w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, w8_t); w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, w9_t); wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, wa_t); wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F1, A, B, C, D, E, wb_t); wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F1, E, A, B, C, D, wc_t); wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F1, D, E, A, B, C, wd_t); we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, C, D, E, A, B, we_t); wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, B, C, D, E, A, wf_t); digest[0] += A; digest[1] += B; digest[2] += C; digest[3] += D; digest[4] += E; } static void hmac_sha1_pad (u32x w0[4], u32x w1[4], u32x w2[4], u32x w3[4], u32x ipad[5], u32x opad[5]) { w0[0] = w0[0] ^ 0x36363636; w0[1] = w0[1] ^ 0x36363636; w0[2] = w0[2] ^ 0x36363636; w0[3] = w0[3] ^ 0x36363636; w1[0] = w1[0] ^ 0x36363636; w1[1] = w1[1] ^ 0x36363636; w1[2] = w1[2] ^ 0x36363636; w1[3] = w1[3] ^ 0x36363636; w2[0] = w2[0] ^ 0x36363636; w2[1] = w2[1] ^ 0x36363636; w2[2] = w2[2] ^ 0x36363636; w2[3] = w2[3] ^ 0x36363636; w3[0] = w3[0] ^ 0x36363636; w3[1] = w3[1] ^ 0x36363636; w3[2] = w3[2] ^ 0x36363636; w3[3] = w3[3] ^ 0x36363636; ipad[0] = SHA1M_A; ipad[1] = SHA1M_B; ipad[2] = SHA1M_C; ipad[3] = SHA1M_D; ipad[4] = SHA1M_E; sha1_transform (w0, w1, w2, w3, ipad); w0[0] = w0[0] ^ 0x6a6a6a6a; w0[1] = w0[1] ^ 0x6a6a6a6a; w0[2] = w0[2] ^ 0x6a6a6a6a; w0[3] = w0[3] ^ 0x6a6a6a6a; w1[0] = w1[0] ^ 0x6a6a6a6a; w1[1] = w1[1] ^ 0x6a6a6a6a; w1[2] = w1[2] ^ 0x6a6a6a6a; w1[3] = w1[3] ^ 0x6a6a6a6a; w2[0] = w2[0] ^ 0x6a6a6a6a; w2[1] = w2[1] ^ 0x6a6a6a6a; w2[2] = w2[2] ^ 0x6a6a6a6a; w2[3] = w2[3] ^ 0x6a6a6a6a; w3[0] = w3[0] ^ 0x6a6a6a6a; w3[1] = w3[1] ^ 0x6a6a6a6a; w3[2] = w3[2] ^ 0x6a6a6a6a; w3[3] = w3[3] ^ 0x6a6a6a6a; opad[0] = SHA1M_A; opad[1] = SHA1M_B; opad[2] = SHA1M_C; opad[3] = SHA1M_D; opad[4] = SHA1M_E; sha1_transform (w0, w1, w2, w3, opad); } static void hmac_sha1_run (u32x w0[4], u32x w1[4], u32x w2[4], u32x w3[4], u32x ipad[5], u32x opad[5], u32x digest[5]) { digest[0] = ipad[0]; digest[1] = ipad[1]; digest[2] = ipad[2]; digest[3] = ipad[3]; digest[4] = ipad[4]; sha1_transform (w0, w1, w2, w3, digest); w0[0] = digest[0]; w0[1] = digest[1]; w0[2] = digest[2]; w0[3] = digest[3]; w1[0] = digest[4]; w1[1] = 0x80000000; w1[2] = 0; w1[3] = 0; w2[0] = 0; w2[1] = 0; w2[2] = 0; w2[3] = 0; w3[0] = 0; w3[1] = 0; w3[2] = 0; w3[3] = (64 + 20) * 8; digest[0] = opad[0]; digest[1] = opad[1]; digest[2] = opad[2]; digest[3] = opad[3]; digest[4] = opad[4]; sha1_transform (w0, w1, w2, w3, digest); } __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m02100_init (__global pw_t *pws, __global gpu_rule_t *rules_buf, __global comb_t *combs_buf, __global bf_t *bfs_buf, __global dcc2_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) { /** * base */ const u32 gid = get_global_id (0); if (gid >= gid_max) return; u32x w0[4]; w0[0] = pws[gid].i[ 0]; w0[1] = pws[gid].i[ 1]; w0[2] = pws[gid].i[ 2]; w0[3] = pws[gid].i[ 3]; u32x w1[4]; w1[0] = 0; w1[1] = 0; w1[2] = 0; w1[3] = 0; u32x w2[4]; w2[0] = 0; w2[1] = 0; w2[2] = 0; w2[3] = 0; u32x w3[4]; w3[0] = 0; w3[1] = 0; w3[2] = 0; w3[3] = 0; const u32 pw_len = pws[gid].pw_len; /** * salt */ u32 salt_len = salt_bufs[salt_pos].salt_len; u32 salt_buf0[4]; u32 salt_buf1[4]; u32 salt_buf2[4]; salt_buf0[0] = salt_bufs[salt_pos].salt_buf[0]; salt_buf0[1] = salt_bufs[salt_pos].salt_buf[1]; salt_buf0[2] = salt_bufs[salt_pos].salt_buf[2]; salt_buf0[3] = salt_bufs[salt_pos].salt_buf[3]; salt_buf1[0] = salt_bufs[salt_pos].salt_buf[4]; salt_buf1[1] = salt_bufs[salt_pos].salt_buf[5]; salt_buf1[2] = salt_bufs[salt_pos].salt_buf[6]; salt_buf1[3] = salt_bufs[salt_pos].salt_buf[7]; salt_buf2[0] = salt_bufs[salt_pos].salt_buf[8]; salt_buf2[1] = salt_bufs[salt_pos].salt_buf[9]; salt_buf2[2] = 0; salt_buf2[3] = 0; /** * generate dcc */ append_0x80_1 (w0, pw_len); make_unicode (w0, w0, w1); w3[2] = pw_len * 2 * 8; u32x digest_md4[4]; digest_md4[0] = MD4M_A; digest_md4[1] = MD4M_B; digest_md4[2] = MD4M_C; digest_md4[3] = MD4M_D; md4_transform (w0, w1, w2, w3, digest_md4); w0[0] = digest_md4[0]; w0[1] = digest_md4[1]; w0[2] = digest_md4[2]; w0[3] = digest_md4[3]; w1[0] = salt_buf0[0]; w1[1] = salt_buf0[1]; w1[2] = salt_buf0[2]; w1[3] = salt_buf0[3]; w2[0] = salt_buf1[0]; w2[1] = salt_buf1[1]; w2[2] = salt_buf1[2]; w2[3] = salt_buf1[3]; w3[0] = salt_buf2[0]; w3[1] = salt_buf2[1]; w3[2] = (16 + salt_len) * 8; w3[3] = 0; append_0x80_4 (w0, w1, w2, w3, 16 + salt_len); digest_md4[0] = MD4M_A; digest_md4[1] = MD4M_B; digest_md4[2] = MD4M_C; digest_md4[3] = MD4M_D; md4_transform (w0, w1, w2, w3, digest_md4); /** * pads */ w0[0] = swap_workaround (digest_md4[0]); w0[1] = swap_workaround (digest_md4[1]); w0[2] = swap_workaround (digest_md4[2]); w0[3] = swap_workaround (digest_md4[3]); w1[0] = 0; w1[1] = 0; w1[2] = 0; w1[3] = 0; w2[0] = 0; w2[1] = 0; w2[2] = 0; w2[3] = 0; w3[0] = 0; w3[1] = 0; w3[2] = 0; w3[3] = 0; u32x ipad[5]; u32x opad[5]; hmac_sha1_pad (w0, w1, w2, w3, ipad, opad); tmps[gid].ipad[0] = ipad[0]; tmps[gid].ipad[1] = ipad[1]; tmps[gid].ipad[2] = ipad[2]; tmps[gid].ipad[3] = ipad[3]; tmps[gid].ipad[4] = ipad[4]; tmps[gid].opad[0] = opad[0]; tmps[gid].opad[1] = opad[1]; tmps[gid].opad[2] = opad[2]; tmps[gid].opad[3] = opad[3]; tmps[gid].opad[4] = opad[4]; /** * hmac1 */ w0[0] = salt_buf0[0]; w0[1] = salt_buf0[1]; w0[2] = salt_buf0[2]; w0[3] = salt_buf0[3]; w1[0] = salt_buf1[0]; w1[1] = salt_buf1[1]; w1[2] = salt_buf1[2]; w1[3] = salt_buf1[3]; w2[0] = salt_buf2[0]; w2[1] = salt_buf2[1]; w2[2] = 0; w2[3] = 0; w3[0] = 0; w3[1] = 0; w3[2] = 0; w3[3] = (64 + salt_len + 4) * 8; append_0x01_4 (w0, w1, w2, w3, salt_len + 3); append_0x80_4 (w0, w1, w2, w3, salt_len + 4); w0[0] = swap_workaround (w0[0]); w0[1] = swap_workaround (w0[1]); w0[2] = swap_workaround (w0[2]); w0[3] = swap_workaround (w0[3]); w1[0] = swap_workaround (w1[0]); w1[1] = swap_workaround (w1[1]); w1[2] = swap_workaround (w1[2]); w1[3] = swap_workaround (w1[3]); w2[0] = swap_workaround (w2[0]); w2[1] = swap_workaround (w2[1]); w2[2] = swap_workaround (w2[2]); w2[3] = swap_workaround (w2[3]); w3[0] = swap_workaround (w3[0]); w3[1] = swap_workaround (w3[1]); u32x digest[5]; hmac_sha1_run (w0, w1, w2, w3, ipad, opad, digest); tmps[gid].dgst[0] = digest[0]; tmps[gid].dgst[1] = digest[1]; tmps[gid].dgst[2] = digest[2]; tmps[gid].dgst[3] = digest[3]; tmps[gid].dgst[4] = digest[4]; tmps[gid].out[0] = digest[0]; tmps[gid].out[1] = digest[1]; tmps[gid].out[2] = digest[2]; tmps[gid].out[3] = digest[3]; } __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m02100_loop (__global pw_t *pws, __global gpu_rule_t *rules_buf, __global comb_t *combs_buf, __global bf_t *bfs_buf, __global dcc2_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) { /** * base */ const u32 gid = get_global_id (0); if (gid >= gid_max) return; u32x ipad[5]; u32x opad[5]; ipad[0] = tmps[gid].ipad[0]; ipad[1] = tmps[gid].ipad[1]; ipad[2] = tmps[gid].ipad[2]; ipad[3] = tmps[gid].ipad[3]; ipad[4] = tmps[gid].ipad[4]; opad[0] = tmps[gid].opad[0]; opad[1] = tmps[gid].opad[1]; opad[2] = tmps[gid].opad[2]; opad[3] = tmps[gid].opad[3]; opad[4] = tmps[gid].opad[4]; /** * iter1 */ u32x dgst[5]; u32x out[4]; dgst[0] = tmps[gid].dgst[0]; dgst[1] = tmps[gid].dgst[1]; dgst[2] = tmps[gid].dgst[2]; dgst[3] = tmps[gid].dgst[3]; dgst[4] = tmps[gid].dgst[4]; out[0] = tmps[gid].out[0]; out[1] = tmps[gid].out[1]; out[2] = tmps[gid].out[2]; out[3] = tmps[gid].out[3]; for (u32 i = 0; i < loop_cnt; i++) { u32x w0[4]; u32x w1[4]; u32x w2[4]; u32x w3[4]; w0[0] = dgst[0]; w0[1] = dgst[1]; w0[2] = dgst[2]; w0[3] = dgst[3]; w1[0] = dgst[4]; w1[1] = 0x80000000; w1[2] = 0; w1[3] = 0; w2[0] = 0; w2[1] = 0; w2[2] = 0; w2[3] = 0; w3[0] = 0; w3[1] = 0; w3[2] = 0; w3[3] = (64 + 20) * 8; hmac_sha1_run (w0, w1, w2, w3, ipad, opad, dgst); out[0] ^= dgst[0]; out[1] ^= dgst[1]; out[2] ^= dgst[2]; out[3] ^= dgst[3]; } tmps[gid].dgst[0] = dgst[0]; tmps[gid].dgst[1] = dgst[1]; tmps[gid].dgst[2] = dgst[2]; tmps[gid].dgst[3] = dgst[3]; tmps[gid].dgst[4] = dgst[4]; tmps[gid].out[0] = out[0]; tmps[gid].out[1] = out[1]; tmps[gid].out[2] = out[2]; tmps[gid].out[3] = out[3]; } __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m02100_comp (__global pw_t *pws, __global gpu_rule_t *rules_buf, __global comb_t *combs_buf, __global bf_t *bfs_buf, __global dcc2_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) { /** * base */ const u32 gid = get_global_id (0); if (gid >= gid_max) return; const u32 lid = get_local_id (0); const u32x r0 = tmps[gid].out[DGST_R0]; const u32x r1 = tmps[gid].out[DGST_R1]; const u32x r2 = tmps[gid].out[DGST_R2]; const u32x r3 = tmps[gid].out[DGST_R3]; #define il_pos 0 #include VECT_COMPARE_M }