/** * Author......: Jens Steube * License.....: MIT */ #define _MD5_ #include "include/constants.h" #include "include/kernel_vendor.h" #define DGST_R0 0 #define DGST_R1 1 #define DGST_R2 2 #define DGST_R3 3 #include "include/kernel_functions.c" #include "OpenCL/types_ocl.c" #include "OpenCL/common.c" #include "include/rp_kernel.h" #include "OpenCL/rp.c" #define COMPARE_S "OpenCL/check_single_comp4.c" #define COMPARE_M "OpenCL/check_multi_comp4.c" __constant u32 padding[8] = { 0x5e4ebf28, 0x418a754e, 0x564e0064, 0x0801faff, 0xb6002e2e, 0x803e68d0, 0xfea90c2f, 0x7a695364 }; static void md5_transform (const u32 w0[4], const u32 w1[4], const u32 w2[4], const u32 w3[4], u32 digest[4]) { u32 a = digest[0]; u32 b = digest[1]; u32 c = digest[2]; u32 d = digest[3]; u32 w0_t = w0[0]; u32 w1_t = w0[1]; u32 w2_t = w0[2]; u32 w3_t = w0[3]; u32 w4_t = w1[0]; u32 w5_t = w1[1]; u32 w6_t = w1[2]; u32 w7_t = w1[3]; u32 w8_t = w2[0]; u32 w9_t = w2[1]; u32 wa_t = w2[2]; u32 wb_t = w2[3]; u32 wc_t = w3[0]; u32 wd_t = w3[1]; u32 we_t = w3[2]; u32 wf_t = w3[3]; MD5_STEP (MD5_Fo, a, b, c, d, w0_t, MD5C00, MD5S00); MD5_STEP (MD5_Fo, d, a, b, c, w1_t, MD5C01, MD5S01); MD5_STEP (MD5_Fo, c, d, a, b, w2_t, MD5C02, MD5S02); MD5_STEP (MD5_Fo, b, c, d, a, w3_t, MD5C03, MD5S03); MD5_STEP (MD5_Fo, a, b, c, d, w4_t, MD5C04, MD5S00); MD5_STEP (MD5_Fo, d, a, b, c, w5_t, MD5C05, MD5S01); MD5_STEP (MD5_Fo, c, d, a, b, w6_t, MD5C06, MD5S02); MD5_STEP (MD5_Fo, b, c, d, a, w7_t, MD5C07, MD5S03); MD5_STEP (MD5_Fo, a, b, c, d, w8_t, MD5C08, MD5S00); MD5_STEP (MD5_Fo, d, a, b, c, w9_t, MD5C09, MD5S01); MD5_STEP (MD5_Fo, c, d, a, b, wa_t, MD5C0a, MD5S02); MD5_STEP (MD5_Fo, b, c, d, a, wb_t, MD5C0b, MD5S03); MD5_STEP (MD5_Fo, a, b, c, d, wc_t, MD5C0c, MD5S00); MD5_STEP (MD5_Fo, d, a, b, c, wd_t, MD5C0d, MD5S01); MD5_STEP (MD5_Fo, c, d, a, b, we_t, MD5C0e, MD5S02); MD5_STEP (MD5_Fo, b, c, d, a, wf_t, MD5C0f, MD5S03); MD5_STEP (MD5_Go, a, b, c, d, w1_t, MD5C10, MD5S10); MD5_STEP (MD5_Go, d, a, b, c, w6_t, MD5C11, MD5S11); MD5_STEP (MD5_Go, c, d, a, b, wb_t, MD5C12, MD5S12); MD5_STEP (MD5_Go, b, c, d, a, w0_t, MD5C13, MD5S13); MD5_STEP (MD5_Go, a, b, c, d, w5_t, MD5C14, MD5S10); MD5_STEP (MD5_Go, d, a, b, c, wa_t, MD5C15, MD5S11); MD5_STEP (MD5_Go, c, d, a, b, wf_t, MD5C16, MD5S12); MD5_STEP (MD5_Go, b, c, d, a, w4_t, MD5C17, MD5S13); MD5_STEP (MD5_Go, a, b, c, d, w9_t, MD5C18, MD5S10); MD5_STEP (MD5_Go, d, a, b, c, we_t, MD5C19, MD5S11); MD5_STEP (MD5_Go, c, d, a, b, w3_t, MD5C1a, MD5S12); MD5_STEP (MD5_Go, b, c, d, a, w8_t, MD5C1b, MD5S13); MD5_STEP (MD5_Go, a, b, c, d, wd_t, MD5C1c, MD5S10); MD5_STEP (MD5_Go, d, a, b, c, w2_t, MD5C1d, MD5S11); MD5_STEP (MD5_Go, c, d, a, b, w7_t, MD5C1e, MD5S12); MD5_STEP (MD5_Go, b, c, d, a, wc_t, MD5C1f, MD5S13); MD5_STEP (MD5_H , a, b, c, d, w5_t, MD5C20, MD5S20); MD5_STEP (MD5_H , d, a, b, c, w8_t, MD5C21, MD5S21); MD5_STEP (MD5_H , c, d, a, b, wb_t, MD5C22, MD5S22); MD5_STEP (MD5_H , b, c, d, a, we_t, MD5C23, MD5S23); MD5_STEP (MD5_H , a, b, c, d, w1_t, MD5C24, MD5S20); MD5_STEP (MD5_H , d, a, b, c, w4_t, MD5C25, MD5S21); MD5_STEP (MD5_H , c, d, a, b, w7_t, MD5C26, MD5S22); MD5_STEP (MD5_H , b, c, d, a, wa_t, MD5C27, MD5S23); MD5_STEP (MD5_H , a, b, c, d, wd_t, MD5C28, MD5S20); MD5_STEP (MD5_H , d, a, b, c, w0_t, MD5C29, MD5S21); MD5_STEP (MD5_H , c, d, a, b, w3_t, MD5C2a, MD5S22); MD5_STEP (MD5_H , b, c, d, a, w6_t, MD5C2b, MD5S23); MD5_STEP (MD5_H , a, b, c, d, w9_t, MD5C2c, MD5S20); MD5_STEP (MD5_H , d, a, b, c, wc_t, MD5C2d, MD5S21); MD5_STEP (MD5_H , c, d, a, b, wf_t, MD5C2e, MD5S22); MD5_STEP (MD5_H , b, c, d, a, w2_t, MD5C2f, MD5S23); MD5_STEP (MD5_I , a, b, c, d, w0_t, MD5C30, MD5S30); MD5_STEP (MD5_I , d, a, b, c, w7_t, MD5C31, MD5S31); MD5_STEP (MD5_I , c, d, a, b, we_t, MD5C32, MD5S32); MD5_STEP (MD5_I , b, c, d, a, w5_t, MD5C33, MD5S33); MD5_STEP (MD5_I , a, b, c, d, wc_t, MD5C34, MD5S30); MD5_STEP (MD5_I , d, a, b, c, w3_t, MD5C35, MD5S31); MD5_STEP (MD5_I , c, d, a, b, wa_t, MD5C36, MD5S32); MD5_STEP (MD5_I , b, c, d, a, w1_t, MD5C37, MD5S33); MD5_STEP (MD5_I , a, b, c, d, w8_t, MD5C38, MD5S30); MD5_STEP (MD5_I , d, a, b, c, wf_t, MD5C39, MD5S31); MD5_STEP (MD5_I , c, d, a, b, w6_t, MD5C3a, MD5S32); MD5_STEP (MD5_I , b, c, d, a, wd_t, MD5C3b, MD5S33); MD5_STEP (MD5_I , a, b, c, d, w4_t, MD5C3c, MD5S30); MD5_STEP (MD5_I , d, a, b, c, wb_t, MD5C3d, MD5S31); MD5_STEP (MD5_I , c, d, a, b, w2_t, MD5C3e, MD5S32); MD5_STEP (MD5_I , b, c, d, a, w9_t, MD5C3f, MD5S33); digest[0] += a; digest[1] += b; digest[2] += c; digest[3] += d; } __kernel void m10420_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, __global comb_t *combs_buf, __global bf_t *bfs_buf, __global void *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 pdf_t *pdf_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) { /** * modifier */ const u32 lid = get_local_id (0); /** * base */ const u32 gid = get_global_id (0); if (gid >= gid_max) return; u32 pw_buf0[4]; pw_buf0[0] = pws[gid].i[ 0]; pw_buf0[1] = pws[gid].i[ 1]; pw_buf0[2] = pws[gid].i[ 2]; pw_buf0[3] = pws[gid].i[ 3]; u32 pw_buf1[4]; pw_buf1[0] = pws[gid].i[ 4]; pw_buf1[1] = pws[gid].i[ 5]; pw_buf1[2] = pws[gid].i[ 6]; pw_buf1[3] = pws[gid].i[ 7]; const u32 pw_len = pws[gid].pw_len; /** * U_buf */ u32 o_buf[8]; o_buf[0] = pdf_bufs[salt_pos].o_buf[0]; o_buf[1] = pdf_bufs[salt_pos].o_buf[1]; o_buf[2] = pdf_bufs[salt_pos].o_buf[2]; o_buf[3] = pdf_bufs[salt_pos].o_buf[3]; o_buf[4] = pdf_bufs[salt_pos].o_buf[4]; o_buf[5] = pdf_bufs[salt_pos].o_buf[5]; o_buf[6] = pdf_bufs[salt_pos].o_buf[6]; o_buf[7] = pdf_bufs[salt_pos].o_buf[7]; u32 P = pdf_bufs[salt_pos].P; u32 id_buf[4]; id_buf[0] = pdf_bufs[salt_pos].id_buf[0]; id_buf[1] = pdf_bufs[salt_pos].id_buf[1]; id_buf[2] = pdf_bufs[salt_pos].id_buf[2]; id_buf[3] = pdf_bufs[salt_pos].id_buf[3]; /** * loop */ for (u32 il_pos = 0; il_pos < rules_cnt; il_pos++) { u32 w0[4]; w0[0] = pw_buf0[0]; w0[1] = pw_buf0[1]; w0[2] = pw_buf0[2]; w0[3] = pw_buf0[3]; u32 w1[4]; w1[0] = pw_buf1[0]; w1[1] = pw_buf1[1]; w1[2] = pw_buf1[2]; w1[3] = pw_buf1[3]; u32 w2[4]; w2[0] = 0; w2[1] = 0; w2[2] = 0; w2[3] = 0; u32 w3[4]; w3[0] = 0; w3[1] = 0; w3[2] = 0; w3[3] = 0; const u32 out_len = apply_rules (rules_buf[il_pos].cmds, w0, w1, pw_len); u32 w0_t[4]; u32 w1_t[4]; u32 w2_t[4]; u32 w3_t[4]; // max length supported by pdf11 is 32 w0_t[0] = padding[0]; w0_t[1] = padding[1]; w0_t[2] = padding[2]; w0_t[3] = padding[3]; w1_t[0] = padding[4]; w1_t[1] = padding[5]; w1_t[2] = padding[6]; w1_t[3] = padding[7]; w2_t[0] = 0; w2_t[1] = 0; w2_t[2] = 0; w2_t[3] = 0; w3_t[0] = 0; w3_t[1] = 0; w3_t[2] = 0; w3_t[3] = 0; switch_buffer_by_offset (w0_t, w1_t, w2_t, w3_t, pw_len); // add password // truncate at 32 is wanted, not a bug! // add o_buf w0_t[0] |= w0[0]; w0_t[1] |= w0[1]; w0_t[2] |= w0[2]; w0_t[3] |= w0[3]; w1_t[0] |= w1[0]; w1_t[1] |= w1[1]; w1_t[2] |= w1[2]; w1_t[3] |= w1[3]; w2_t[0] = o_buf[0]; w2_t[1] = o_buf[1]; w2_t[2] = o_buf[2]; w2_t[3] = o_buf[3]; w3_t[0] = o_buf[4]; w3_t[1] = o_buf[5]; w3_t[2] = o_buf[6]; w3_t[3] = o_buf[7]; u32 digest[4]; digest[0] = MD5M_A; digest[1] = MD5M_B; digest[2] = MD5M_C; digest[3] = MD5M_D; md5_transform (w0_t, w1_t, w2_t, w3_t, digest); w0_t[0] = P; w0_t[1] = id_buf[0]; w0_t[2] = id_buf[1]; w0_t[3] = id_buf[2]; w1_t[0] = id_buf[3]; w1_t[1] = 0x80; w1_t[2] = 0; w1_t[3] = 0; w2_t[0] = 0; w2_t[1] = 0; w2_t[2] = 0; w2_t[3] = 0; w3_t[0] = 0; w3_t[1] = 0; w3_t[2] = 84 * 8; w3_t[3] = 0; md5_transform (w0_t, w1_t, w2_t, w3_t, digest); u32 a = digest[0]; u32 b = digest[1] & 0xff; const u32 r0 = a; const u32 r1 = b; const u32 r2 = 0; const u32 r3 = 0; #include COMPARE_M } } __kernel void m10420_m08 (__global pw_t *pws, __global kernel_rule_t *rules_buf, __global comb_t *combs_buf, __global bf_t *bfs_buf, __global void *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 pdf_t *pdf_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) { } __kernel void m10420_m16 (__global pw_t *pws, __global kernel_rule_t *rules_buf, __global comb_t *combs_buf, __global bf_t *bfs_buf, __global void *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 pdf_t *pdf_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) { } __kernel void m10420_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, __global comb_t *combs_buf, __global bf_t *bfs_buf, __global void *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 pdf_t *pdf_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) { /** * modifier */ const u32 lid = get_local_id (0); /** * base */ const u32 gid = get_global_id (0); if (gid >= gid_max) return; u32 pw_buf0[4]; pw_buf0[0] = pws[gid].i[ 0]; pw_buf0[1] = pws[gid].i[ 1]; pw_buf0[2] = pws[gid].i[ 2]; pw_buf0[3] = pws[gid].i[ 3]; u32 pw_buf1[4]; pw_buf1[0] = pws[gid].i[ 4]; pw_buf1[1] = pws[gid].i[ 5]; pw_buf1[2] = pws[gid].i[ 6]; pw_buf1[3] = pws[gid].i[ 7]; const u32 pw_len = pws[gid].pw_len; /** * digest */ const u32 search[4] = { digests_buf[digests_offset].digest_buf[DGST_R0], digests_buf[digests_offset].digest_buf[DGST_R1], digests_buf[digests_offset].digest_buf[DGST_R2], digests_buf[digests_offset].digest_buf[DGST_R3] }; /** * U_buf */ u32 o_buf[8]; o_buf[0] = pdf_bufs[salt_pos].o_buf[0]; o_buf[1] = pdf_bufs[salt_pos].o_buf[1]; o_buf[2] = pdf_bufs[salt_pos].o_buf[2]; o_buf[3] = pdf_bufs[salt_pos].o_buf[3]; o_buf[4] = pdf_bufs[salt_pos].o_buf[4]; o_buf[5] = pdf_bufs[salt_pos].o_buf[5]; o_buf[6] = pdf_bufs[salt_pos].o_buf[6]; o_buf[7] = pdf_bufs[salt_pos].o_buf[7]; u32 P = pdf_bufs[salt_pos].P; u32 id_buf[4]; id_buf[0] = pdf_bufs[salt_pos].id_buf[0]; id_buf[1] = pdf_bufs[salt_pos].id_buf[1]; id_buf[2] = pdf_bufs[salt_pos].id_buf[2]; id_buf[3] = pdf_bufs[salt_pos].id_buf[3]; /** * loop */ for (u32 il_pos = 0; il_pos < rules_cnt; il_pos++) { u32 w0[4]; w0[0] = pw_buf0[0]; w0[1] = pw_buf0[1]; w0[2] = pw_buf0[2]; w0[3] = pw_buf0[3]; u32 w1[4]; w1[0] = pw_buf1[0]; w1[1] = pw_buf1[1]; w1[2] = pw_buf1[2]; w1[3] = pw_buf1[3]; u32 w2[4]; w2[0] = 0; w2[1] = 0; w2[2] = 0; w2[3] = 0; u32 w3[4]; w3[0] = 0; w3[1] = 0; w3[2] = 0; w3[3] = 0; const u32 out_len = apply_rules (rules_buf[il_pos].cmds, w0, w1, pw_len); u32 w0_t[4]; u32 w1_t[4]; u32 w2_t[4]; u32 w3_t[4]; // max length supported by pdf11 is 32 w0_t[0] = padding[0]; w0_t[1] = padding[1]; w0_t[2] = padding[2]; w0_t[3] = padding[3]; w1_t[0] = padding[4]; w1_t[1] = padding[5]; w1_t[2] = padding[6]; w1_t[3] = padding[7]; w2_t[0] = 0; w2_t[1] = 0; w2_t[2] = 0; w2_t[3] = 0; w3_t[0] = 0; w3_t[1] = 0; w3_t[2] = 0; w3_t[3] = 0; switch_buffer_by_offset (w0_t, w1_t, w2_t, w3_t, pw_len); // add password // truncate at 32 is wanted, not a bug! // add o_buf w0_t[0] |= w0[0]; w0_t[1] |= w0[1]; w0_t[2] |= w0[2]; w0_t[3] |= w0[3]; w1_t[0] |= w1[0]; w1_t[1] |= w1[1]; w1_t[2] |= w1[2]; w1_t[3] |= w1[3]; w2_t[0] = o_buf[0]; w2_t[1] = o_buf[1]; w2_t[2] = o_buf[2]; w2_t[3] = o_buf[3]; w3_t[0] = o_buf[4]; w3_t[1] = o_buf[5]; w3_t[2] = o_buf[6]; w3_t[3] = o_buf[7]; u32 digest[4]; digest[0] = MD5M_A; digest[1] = MD5M_B; digest[2] = MD5M_C; digest[3] = MD5M_D; md5_transform (w0_t, w1_t, w2_t, w3_t, digest); w0_t[0] = P; w0_t[1] = id_buf[0]; w0_t[2] = id_buf[1]; w0_t[3] = id_buf[2]; w1_t[0] = id_buf[3]; w1_t[1] = 0x80; w1_t[2] = 0; w1_t[3] = 0; w2_t[0] = 0; w2_t[1] = 0; w2_t[2] = 0; w2_t[3] = 0; w3_t[0] = 0; w3_t[1] = 0; w3_t[2] = 84 * 8; w3_t[3] = 0; md5_transform (w0_t, w1_t, w2_t, w3_t, digest); u32 a = digest[0]; u32 b = digest[1] & 0xff; const u32 r0 = a; const u32 r1 = b; const u32 r2 = 0; const u32 r3 = 0; #include COMPARE_S } } __kernel void m10420_s08 (__global pw_t *pws, __global kernel_rule_t *rules_buf, __global comb_t *combs_buf, __global bf_t *bfs_buf, __global void *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 pdf_t *pdf_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) { } __kernel void m10420_s16 (__global pw_t *pws, __global kernel_rule_t *rules_buf, __global comb_t *combs_buf, __global bf_t *bfs_buf, __global void *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 pdf_t *pdf_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) { }