/** * Author......: Jens Steube * License.....: MIT */ #define _SHA512_ #include "include/constants.h" #include "include/kernel_vendor.h" #define DGST_R0 14 #define DGST_R1 15 #define DGST_R2 6 #define DGST_R3 7 #include "include/kernel_functions.c" #include "types_ocl.c" #include "common.c" #define COMPARE_S "check_single_comp4.c" #define COMPARE_M "check_multi_comp4.c" __constant u64 k_sha512[80] = { SHA512C00, SHA512C01, SHA512C02, SHA512C03, SHA512C04, SHA512C05, SHA512C06, SHA512C07, SHA512C08, SHA512C09, SHA512C0a, SHA512C0b, SHA512C0c, SHA512C0d, SHA512C0e, SHA512C0f, SHA512C10, SHA512C11, SHA512C12, SHA512C13, SHA512C14, SHA512C15, SHA512C16, SHA512C17, SHA512C18, SHA512C19, SHA512C1a, SHA512C1b, SHA512C1c, SHA512C1d, SHA512C1e, SHA512C1f, SHA512C20, SHA512C21, SHA512C22, SHA512C23, SHA512C24, SHA512C25, SHA512C26, SHA512C27, SHA512C28, SHA512C29, SHA512C2a, SHA512C2b, SHA512C2c, SHA512C2d, SHA512C2e, SHA512C2f, SHA512C30, SHA512C31, SHA512C32, SHA512C33, SHA512C34, SHA512C35, SHA512C36, SHA512C37, SHA512C38, SHA512C39, SHA512C3a, SHA512C3b, SHA512C3c, SHA512C3d, SHA512C3e, SHA512C3f, SHA512C40, SHA512C41, SHA512C42, SHA512C43, SHA512C44, SHA512C45, SHA512C46, SHA512C47, SHA512C48, SHA512C49, SHA512C4a, SHA512C4b, SHA512C4c, SHA512C4d, SHA512C4e, SHA512C4f, }; static void sha512_transform (const u32 w0[4], const u32 w1[4], const u32 w2[4], const u32 w3[4], u64 digest[8]) { u64 w0_t = hl32_to_64 (w0[0], w0[1]); u64 w1_t = hl32_to_64 (w0[2], w0[3]); u64 w2_t = hl32_to_64 (w1[0], w1[1]); u64 w3_t = hl32_to_64 (w1[2], w1[3]); u64 w4_t = hl32_to_64 (w2[0], w2[1]); u64 w5_t = hl32_to_64 (w2[2], w2[3]); u64 w6_t = hl32_to_64 (w3[0], w3[1]); u64 w7_t = 0; u64 w8_t = 0; u64 w9_t = 0; u64 wa_t = 0; u64 wb_t = 0; u64 wc_t = 0; u64 wd_t = 0; u64 we_t = 0; u64 wf_t = hl32_to_64 (w3[2], w3[3]); u64 a = digest[0]; u64 b = digest[1]; u64 c = digest[2]; u64 d = digest[3]; u64 e = digest[4]; u64 f = digest[5]; u64 g = digest[6]; u64 h = digest[7]; #define ROUND_EXPAND() \ { \ w0_t = SHA512_EXPAND (we_t, w9_t, w1_t, w0_t); \ w1_t = SHA512_EXPAND (wf_t, wa_t, w2_t, w1_t); \ w2_t = SHA512_EXPAND (w0_t, wb_t, w3_t, w2_t); \ w3_t = SHA512_EXPAND (w1_t, wc_t, w4_t, w3_t); \ w4_t = SHA512_EXPAND (w2_t, wd_t, w5_t, w4_t); \ w5_t = SHA512_EXPAND (w3_t, we_t, w6_t, w5_t); \ w6_t = SHA512_EXPAND (w4_t, wf_t, w7_t, w6_t); \ w7_t = SHA512_EXPAND (w5_t, w0_t, w8_t, w7_t); \ w8_t = SHA512_EXPAND (w6_t, w1_t, w9_t, w8_t); \ w9_t = SHA512_EXPAND (w7_t, w2_t, wa_t, w9_t); \ wa_t = SHA512_EXPAND (w8_t, w3_t, wb_t, wa_t); \ wb_t = SHA512_EXPAND (w9_t, w4_t, wc_t, wb_t); \ wc_t = SHA512_EXPAND (wa_t, w5_t, wd_t, wc_t); \ wd_t = SHA512_EXPAND (wb_t, w6_t, we_t, wd_t); \ we_t = SHA512_EXPAND (wc_t, w7_t, wf_t, we_t); \ wf_t = SHA512_EXPAND (wd_t, w8_t, w0_t, wf_t); \ } #define ROUND_STEP(i) \ { \ SHA512_STEP (SHA512_F0o, SHA512_F1o, a, b, c, d, e, f, g, h, w0_t, k_sha512[i + 0]); \ SHA512_STEP (SHA512_F0o, SHA512_F1o, h, a, b, c, d, e, f, g, w1_t, k_sha512[i + 1]); \ SHA512_STEP (SHA512_F0o, SHA512_F1o, g, h, a, b, c, d, e, f, w2_t, k_sha512[i + 2]); \ SHA512_STEP (SHA512_F0o, SHA512_F1o, f, g, h, a, b, c, d, e, w3_t, k_sha512[i + 3]); \ SHA512_STEP (SHA512_F0o, SHA512_F1o, e, f, g, h, a, b, c, d, w4_t, k_sha512[i + 4]); \ SHA512_STEP (SHA512_F0o, SHA512_F1o, d, e, f, g, h, a, b, c, w5_t, k_sha512[i + 5]); \ SHA512_STEP (SHA512_F0o, SHA512_F1o, c, d, e, f, g, h, a, b, w6_t, k_sha512[i + 6]); \ SHA512_STEP (SHA512_F0o, SHA512_F1o, b, c, d, e, f, g, h, a, w7_t, k_sha512[i + 7]); \ SHA512_STEP (SHA512_F0o, SHA512_F1o, a, b, c, d, e, f, g, h, w8_t, k_sha512[i + 8]); \ SHA512_STEP (SHA512_F0o, SHA512_F1o, h, a, b, c, d, e, f, g, w9_t, k_sha512[i + 9]); \ SHA512_STEP (SHA512_F0o, SHA512_F1o, g, h, a, b, c, d, e, f, wa_t, k_sha512[i + 10]); \ SHA512_STEP (SHA512_F0o, SHA512_F1o, f, g, h, a, b, c, d, e, wb_t, k_sha512[i + 11]); \ SHA512_STEP (SHA512_F0o, SHA512_F1o, e, f, g, h, a, b, c, d, wc_t, k_sha512[i + 12]); \ SHA512_STEP (SHA512_F0o, SHA512_F1o, d, e, f, g, h, a, b, c, wd_t, k_sha512[i + 13]); \ SHA512_STEP (SHA512_F0o, SHA512_F1o, c, d, e, f, g, h, a, b, we_t, k_sha512[i + 14]); \ SHA512_STEP (SHA512_F0o, SHA512_F1o, b, c, d, e, f, g, h, a, wf_t, k_sha512[i + 15]); \ } ROUND_STEP (0); #pragma unroll for (int i = 16; i < 80; i += 16) { ROUND_EXPAND (); ROUND_STEP (i); } /* rev digest[0] += a; digest[1] += b; digest[2] += c; digest[3] += d; digest[4] += e; digest[5] += f; digest[6] += g; digest[7] += h; */ digest[0] = a; digest[1] = b; digest[2] = c; digest[3] = d; digest[4] = e; digest[5] = f; digest[6] = g; digest[7] = h; } __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m01700_m04 (__global pw_t *pws, __global gpu_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 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 combs_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 wordl0[4]; wordl0[0] = pws[gid].i[ 0]; wordl0[1] = pws[gid].i[ 1]; wordl0[2] = pws[gid].i[ 2]; wordl0[3] = pws[gid].i[ 3]; u32 wordl1[4]; wordl1[0] = pws[gid].i[ 4]; wordl1[1] = pws[gid].i[ 5]; wordl1[2] = pws[gid].i[ 6]; wordl1[3] = pws[gid].i[ 7]; u32 wordl2[4]; wordl2[0] = 0; wordl2[1] = 0; wordl2[2] = 0; wordl2[3] = 0; u32 wordl3[4]; wordl3[0] = 0; wordl3[1] = 0; wordl3[2] = 0; wordl3[3] = 0; const u32 pw_l_len = pws[gid].pw_len; if (combs_mode == COMBINATOR_MODE_BASE_RIGHT) { append_0x80_2x4 (wordl0, wordl1, pw_l_len); switch_buffer_by_offset (wordl0, wordl1, wordl2, wordl3, combs_buf[0].pw_len); } /** * loop */ for (u32 il_pos = 0; il_pos < combs_cnt; il_pos++) { const u32 pw_r_len = combs_buf[il_pos].pw_len; const u32 pw_len = pw_l_len + pw_r_len; u32 wordr0[4]; wordr0[0] = combs_buf[il_pos].i[0]; wordr0[1] = combs_buf[il_pos].i[1]; wordr0[2] = combs_buf[il_pos].i[2]; wordr0[3] = combs_buf[il_pos].i[3]; u32 wordr1[4]; wordr1[0] = combs_buf[il_pos].i[4]; wordr1[1] = combs_buf[il_pos].i[5]; wordr1[2] = combs_buf[il_pos].i[6]; wordr1[3] = combs_buf[il_pos].i[7]; u32 wordr2[4]; wordr2[0] = 0; wordr2[1] = 0; wordr2[2] = 0; wordr2[3] = 0; u32 wordr3[4]; wordr3[0] = 0; wordr3[1] = 0; wordr3[2] = 0; wordr3[3] = 0; if (combs_mode == COMBINATOR_MODE_BASE_LEFT) { append_0x80_2x4 (wordr0, wordr1, pw_r_len); switch_buffer_by_offset (wordr0, wordr1, wordr2, wordr3, pw_l_len); } u32 w0[4]; u32 w1[4]; u32 w2[4]; u32 w3[4]; w0[0] = wordl0[0] | wordr0[0]; w0[1] = wordl0[1] | wordr0[1]; w0[2] = wordl0[2] | wordr0[2]; w0[3] = wordl0[3] | wordr0[3]; w1[0] = wordl1[0] | wordr1[0]; w1[1] = wordl1[1] | wordr1[1]; w1[2] = wordl1[2] | wordr1[2]; w1[3] = wordl1[3] | wordr1[3]; w2[0] = wordl2[0] | wordr2[0]; w2[1] = wordl2[1] | wordr2[1]; w2[2] = wordl2[2] | wordr2[2]; w2[3] = wordl2[3] | wordr2[3]; w3[0] = wordl3[0] | wordr3[0]; w3[1] = wordl3[1] | wordr3[1]; w3[2] = wordl3[2] | wordr3[2]; w3[3] = wordl3[3] | wordr3[3]; /** * SHA512 */ u32 w0_t[4]; u32 w1_t[4]; u32 w2_t[4]; u32 w3_t[4]; w0_t[0] = swap_workaround (w0[0]); w0_t[1] = swap_workaround (w0[1]); w0_t[2] = swap_workaround (w0[2]); w0_t[3] = swap_workaround (w0[3]); w1_t[0] = swap_workaround (w1[0]); w1_t[1] = swap_workaround (w1[1]); w1_t[2] = swap_workaround (w1[2]); w1_t[3] = swap_workaround (w1[3]); w2_t[0] = swap_workaround (w2[0]); w2_t[1] = swap_workaround (w2[1]); w2_t[2] = swap_workaround (w2[2]); w2_t[3] = swap_workaround (w2[3]); w3_t[0] = swap_workaround (w3[0]); w3_t[1] = swap_workaround (w3[1]); w3_t[2] = 0; w3_t[3] = pw_len * 8; u64 digest[8]; digest[0] = SHA512M_A; digest[1] = SHA512M_B; digest[2] = SHA512M_C; digest[3] = SHA512M_D; digest[4] = SHA512M_E; digest[5] = SHA512M_F; digest[6] = SHA512M_G; digest[7] = SHA512M_H; sha512_transform (w0_t, w1_t, w2_t, w3_t, digest); const u32 r0 = l32_from_64 (digest[7]); const u32 r1 = h32_from_64 (digest[7]); const u32 r2 = l32_from_64 (digest[3]); const u32 r3 = h32_from_64 (digest[3]); #include COMPARE_M } } __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m01700_m08 (__global pw_t *pws, __global gpu_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 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 combs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max) { } __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m01700_m16 (__global pw_t *pws, __global gpu_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 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 combs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max) { } __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m01700_s04 (__global pw_t *pws, __global gpu_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 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 combs_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 wordl0[4]; wordl0[0] = pws[gid].i[ 0]; wordl0[1] = pws[gid].i[ 1]; wordl0[2] = pws[gid].i[ 2]; wordl0[3] = pws[gid].i[ 3]; u32 wordl1[4]; wordl1[0] = pws[gid].i[ 4]; wordl1[1] = pws[gid].i[ 5]; wordl1[2] = pws[gid].i[ 6]; wordl1[3] = pws[gid].i[ 7]; u32 wordl2[4]; wordl2[0] = 0; wordl2[1] = 0; wordl2[2] = 0; wordl2[3] = 0; u32 wordl3[4]; wordl3[0] = 0; wordl3[1] = 0; wordl3[2] = 0; wordl3[3] = 0; const u32 pw_l_len = pws[gid].pw_len; if (combs_mode == COMBINATOR_MODE_BASE_RIGHT) { append_0x80_2x4 (wordl0, wordl1, pw_l_len); switch_buffer_by_offset (wordl0, wordl1, wordl2, wordl3, combs_buf[0].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] }; /** * loop */ for (u32 il_pos = 0; il_pos < combs_cnt; il_pos++) { const u32 pw_r_len = combs_buf[il_pos].pw_len; const u32 pw_len = pw_l_len + pw_r_len; u32 wordr0[4]; wordr0[0] = combs_buf[il_pos].i[0]; wordr0[1] = combs_buf[il_pos].i[1]; wordr0[2] = combs_buf[il_pos].i[2]; wordr0[3] = combs_buf[il_pos].i[3]; u32 wordr1[4]; wordr1[0] = combs_buf[il_pos].i[4]; wordr1[1] = combs_buf[il_pos].i[5]; wordr1[2] = combs_buf[il_pos].i[6]; wordr1[3] = combs_buf[il_pos].i[7]; u32 wordr2[4]; wordr2[0] = 0; wordr2[1] = 0; wordr2[2] = 0; wordr2[3] = 0; u32 wordr3[4]; wordr3[0] = 0; wordr3[1] = 0; wordr3[2] = 0; wordr3[3] = 0; if (combs_mode == COMBINATOR_MODE_BASE_LEFT) { append_0x80_2x4 (wordr0, wordr1, pw_r_len); switch_buffer_by_offset (wordr0, wordr1, wordr2, wordr3, pw_l_len); } u32 w0[4]; u32 w1[4]; u32 w2[4]; u32 w3[4]; w0[0] = wordl0[0] | wordr0[0]; w0[1] = wordl0[1] | wordr0[1]; w0[2] = wordl0[2] | wordr0[2]; w0[3] = wordl0[3] | wordr0[3]; w1[0] = wordl1[0] | wordr1[0]; w1[1] = wordl1[1] | wordr1[1]; w1[2] = wordl1[2] | wordr1[2]; w1[3] = wordl1[3] | wordr1[3]; w2[0] = wordl2[0] | wordr2[0]; w2[1] = wordl2[1] | wordr2[1]; w2[2] = wordl2[2] | wordr2[2]; w2[3] = wordl2[3] | wordr2[3]; w3[0] = wordl3[0] | wordr3[0]; w3[1] = wordl3[1] | wordr3[1]; w3[2] = wordl3[2] | wordr3[2]; w3[3] = wordl3[3] | wordr3[3]; /** * SHA512 */ u32 w0_t[4]; u32 w1_t[4]; u32 w2_t[4]; u32 w3_t[4]; w0_t[0] = swap_workaround (w0[0]); w0_t[1] = swap_workaround (w0[1]); w0_t[2] = swap_workaround (w0[2]); w0_t[3] = swap_workaround (w0[3]); w1_t[0] = swap_workaround (w1[0]); w1_t[1] = swap_workaround (w1[1]); w1_t[2] = swap_workaround (w1[2]); w1_t[3] = swap_workaround (w1[3]); w2_t[0] = swap_workaround (w2[0]); w2_t[1] = swap_workaround (w2[1]); w2_t[2] = swap_workaround (w2[2]); w2_t[3] = swap_workaround (w2[3]); w3_t[0] = swap_workaround (w3[0]); w3_t[1] = swap_workaround (w3[1]); w3_t[2] = 0; w3_t[3] = pw_len * 8; u64 digest[8]; digest[0] = SHA512M_A; digest[1] = SHA512M_B; digest[2] = SHA512M_C; digest[3] = SHA512M_D; digest[4] = SHA512M_E; digest[5] = SHA512M_F; digest[6] = SHA512M_G; digest[7] = SHA512M_H; sha512_transform (w0_t, w1_t, w2_t, w3_t, digest); const u32 r0 = l32_from_64 (digest[7]); const u32 r1 = h32_from_64 (digest[7]); const u32 r2 = l32_from_64 (digest[3]); const u32 r3 = h32_from_64 (digest[3]); #include COMPARE_S } } __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m01700_s08 (__global pw_t *pws, __global gpu_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 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 combs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max) { } __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m01700_s16 (__global pw_t *pws, __global gpu_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 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 combs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max) { }