/** * Author......: See docs/credits.txt * License.....: MIT */ #define NEW_SIMD_CODE #include "inc_vendor.cl" #include "inc_hash_constants.h" #include "inc_hash_functions.cl" #include "inc_types.cl" #include "inc_common.cl" #include "inc_simd.cl" #include "inc_hash_sha512.cl" #include "inc_cipher_aes.cl" #include "inc_cipher_twofish.cl" #include "inc_cipher_serpent.cl" #include "inc_truecrypt_crc32.cl" #include "inc_truecrypt_xts.cl" u32 u8add (const u32 a, const u32 b) { const u32 a1 = (a >> 0) & 0xff; const u32 a2 = (a >> 8) & 0xff; const u32 a3 = (a >> 16) & 0xff; const u32 a4 = (a >> 24) & 0xff; const u32 b1 = (b >> 0) & 0xff; const u32 b2 = (b >> 8) & 0xff; const u32 b3 = (b >> 16) & 0xff; const u32 b4 = (b >> 24) & 0xff; const u32 r1 = (a1 + b1) & 0xff; const u32 r2 = (a2 + b2) & 0xff; const u32 r3 = (a3 + b3) & 0xff; const u32 r4 = (a4 + b4) & 0xff; const u32 r = r1 << 0 | r2 << 8 | r3 << 16 | r4 << 24; return r; } void hmac_sha512_run_V (u32x w0[4], u32x w1[4], u32x w2[4], u32x w3[4], u32x w4[4], u32x w5[4], u32x w6[4], u32x w7[4], u64x ipad[8], u64x opad[8], u64x digest[8]) { digest[0] = ipad[0]; digest[1] = ipad[1]; digest[2] = ipad[2]; digest[3] = ipad[3]; digest[4] = ipad[4]; digest[5] = ipad[5]; digest[6] = ipad[6]; digest[7] = ipad[7]; sha512_transform_vector (w0, w1, w2, w3, w4, w5, w6, w7, digest); w0[0] = h32_from_64 (digest[0]); w0[1] = l32_from_64 (digest[0]); w0[2] = h32_from_64 (digest[1]); w0[3] = l32_from_64 (digest[1]); w1[0] = h32_from_64 (digest[2]); w1[1] = l32_from_64 (digest[2]); w1[2] = h32_from_64 (digest[3]); w1[3] = l32_from_64 (digest[3]); w2[0] = h32_from_64 (digest[4]); w2[1] = l32_from_64 (digest[4]); w2[2] = h32_from_64 (digest[5]); w2[3] = l32_from_64 (digest[5]); w3[0] = h32_from_64 (digest[6]); w3[1] = l32_from_64 (digest[6]); w3[2] = h32_from_64 (digest[7]); w3[3] = l32_from_64 (digest[7]); w4[0] = 0x80000000; w4[1] = 0; w4[2] = 0; w4[3] = 0; w5[0] = 0; w5[1] = 0; w5[2] = 0; w5[3] = 0; w6[0] = 0; w6[1] = 0; w6[2] = 0; w6[3] = 0; w7[0] = 0; w7[1] = 0; w7[2] = 0; w7[3] = (128 + 64) * 8; digest[0] = opad[0]; digest[1] = opad[1]; digest[2] = opad[2]; digest[3] = opad[3]; digest[4] = opad[4]; digest[5] = opad[5]; digest[6] = opad[6]; digest[7] = opad[7]; sha512_transform_vector (w0, w1, w2, w3, w4, w5, w6, w7, digest); } __kernel void m06221_init (__global pw_t *pws, __constant const kernel_rule_t *rules_buf, __global const pw_t *combs_buf, __global const bf_t *bfs_buf, __global tc64_tmp_t *tmps, __global void *hooks, __global const u32 *bitmaps_buf_s1_a, __global const u32 *bitmaps_buf_s1_b, __global const u32 *bitmaps_buf_s1_c, __global const u32 *bitmaps_buf_s1_d, __global const u32 *bitmaps_buf_s2_a, __global const u32 *bitmaps_buf_s2_b, __global const u32 *bitmaps_buf_s2_c, __global const u32 *bitmaps_buf_s2_d, __global plain_t *plains_buf, __global const digest_t *digests_buf, __global u32 *hashes_shown, __global const salt_t *salt_bufs, __global const tc_t *esalt_bufs, __global u32 *d_return_buf, __global u32 *d_scryptV0_buf, __global u32 *d_scryptV1_buf, __global u32 *d_scryptV2_buf, __global u32 *d_scryptV3_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 u64 gid_max) { /** * base */ const u64 gid = get_global_id (0); if (gid >= gid_max) return; u32 w0[4]; u32 w1[4]; u32 w2[4]; u32 w3[4]; u32 w4[4]; u32 w5[4]; u32 w6[4]; u32 w7[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]; w1[0] = pws[gid].i[ 4]; w1[1] = pws[gid].i[ 5]; w1[2] = pws[gid].i[ 6]; w1[3] = pws[gid].i[ 7]; w2[0] = pws[gid].i[ 8]; w2[1] = pws[gid].i[ 9]; w2[2] = pws[gid].i[10]; w2[3] = pws[gid].i[11]; w3[0] = pws[gid].i[12]; w3[1] = pws[gid].i[13]; w3[2] = pws[gid].i[14]; w3[3] = pws[gid].i[15]; w4[0] = pws[gid].i[16]; w4[1] = pws[gid].i[17]; w4[2] = pws[gid].i[18]; w4[3] = pws[gid].i[19]; w5[0] = pws[gid].i[20]; w5[1] = pws[gid].i[21]; w5[2] = pws[gid].i[22]; w5[3] = pws[gid].i[23]; w6[0] = pws[gid].i[24]; w6[1] = pws[gid].i[25]; w6[2] = pws[gid].i[26]; w6[3] = pws[gid].i[27]; w7[0] = pws[gid].i[28]; w7[1] = pws[gid].i[29]; w7[2] = pws[gid].i[30]; w7[3] = pws[gid].i[31]; w0[0] = u8add (w0[0], esalt_bufs[digests_offset].keyfile_buf[ 0]); w0[1] = u8add (w0[1], esalt_bufs[digests_offset].keyfile_buf[ 1]); w0[2] = u8add (w0[2], esalt_bufs[digests_offset].keyfile_buf[ 2]); w0[3] = u8add (w0[3], esalt_bufs[digests_offset].keyfile_buf[ 3]); w1[0] = u8add (w1[0], esalt_bufs[digests_offset].keyfile_buf[ 4]); w1[1] = u8add (w1[1], esalt_bufs[digests_offset].keyfile_buf[ 5]); w1[2] = u8add (w1[2], esalt_bufs[digests_offset].keyfile_buf[ 6]); w1[3] = u8add (w1[3], esalt_bufs[digests_offset].keyfile_buf[ 7]); w2[0] = u8add (w2[0], esalt_bufs[digests_offset].keyfile_buf[ 8]); w2[1] = u8add (w2[1], esalt_bufs[digests_offset].keyfile_buf[ 9]); w2[2] = u8add (w2[2], esalt_bufs[digests_offset].keyfile_buf[10]); w2[3] = u8add (w2[3], esalt_bufs[digests_offset].keyfile_buf[11]); w3[0] = u8add (w3[0], esalt_bufs[digests_offset].keyfile_buf[12]); w3[1] = u8add (w3[1], esalt_bufs[digests_offset].keyfile_buf[13]); w3[2] = u8add (w3[2], esalt_bufs[digests_offset].keyfile_buf[14]); w3[3] = u8add (w3[3], esalt_bufs[digests_offset].keyfile_buf[15]); w0[0] = swap32_S (w0[0]); w0[1] = swap32_S (w0[1]); w0[2] = swap32_S (w0[2]); w0[3] = swap32_S (w0[3]); w1[0] = swap32_S (w1[0]); w1[1] = swap32_S (w1[1]); w1[2] = swap32_S (w1[2]); w1[3] = swap32_S (w1[3]); w2[0] = swap32_S (w2[0]); w2[1] = swap32_S (w2[1]); w2[2] = swap32_S (w2[2]); w2[3] = swap32_S (w2[3]); w3[0] = swap32_S (w3[0]); w3[1] = swap32_S (w3[1]); w3[2] = swap32_S (w3[2]); w3[3] = swap32_S (w3[3]); w4[0] = swap32_S (w4[0]); w4[1] = swap32_S (w4[1]); w4[2] = swap32_S (w4[2]); w4[3] = swap32_S (w4[3]); w5[0] = swap32_S (w5[0]); w5[1] = swap32_S (w5[1]); w5[2] = swap32_S (w5[2]); w5[3] = swap32_S (w5[3]); w6[0] = swap32_S (w6[0]); w6[1] = swap32_S (w6[1]); w6[2] = swap32_S (w6[2]); w6[3] = swap32_S (w6[3]); w7[0] = swap32_S (w7[0]); w7[1] = swap32_S (w7[1]); w7[2] = swap32_S (w7[2]); w7[3] = swap32_S (w7[3]); sha512_hmac_ctx_t sha512_hmac_ctx; sha512_hmac_init_128 (&sha512_hmac_ctx, w0, w1, w2, w3, w5, w5, w6, w7); tmps[gid].ipad[0] = sha512_hmac_ctx.ipad.h[0]; tmps[gid].ipad[1] = sha512_hmac_ctx.ipad.h[1]; tmps[gid].ipad[2] = sha512_hmac_ctx.ipad.h[2]; tmps[gid].ipad[3] = sha512_hmac_ctx.ipad.h[3]; tmps[gid].ipad[4] = sha512_hmac_ctx.ipad.h[4]; tmps[gid].ipad[5] = sha512_hmac_ctx.ipad.h[5]; tmps[gid].ipad[6] = sha512_hmac_ctx.ipad.h[6]; tmps[gid].ipad[7] = sha512_hmac_ctx.ipad.h[7]; tmps[gid].opad[0] = sha512_hmac_ctx.opad.h[0]; tmps[gid].opad[1] = sha512_hmac_ctx.opad.h[1]; tmps[gid].opad[2] = sha512_hmac_ctx.opad.h[2]; tmps[gid].opad[3] = sha512_hmac_ctx.opad.h[3]; tmps[gid].opad[4] = sha512_hmac_ctx.opad.h[4]; tmps[gid].opad[5] = sha512_hmac_ctx.opad.h[5]; tmps[gid].opad[6] = sha512_hmac_ctx.opad.h[6]; tmps[gid].opad[7] = sha512_hmac_ctx.opad.h[7]; sha512_hmac_update_global_swap (&sha512_hmac_ctx, esalt_bufs[digests_offset].salt_buf, 64); for (u32 i = 0, j = 1; i < 8; i += 8, j += 1) { sha512_hmac_ctx_t sha512_hmac_ctx2 = sha512_hmac_ctx; w0[0] = j; w0[1] = 0; w0[2] = 0; w0[3] = 0; 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; w4[0] = 0; w4[1] = 0; w4[2] = 0; w4[3] = 0; w5[0] = 0; w5[1] = 0; w5[2] = 0; w5[3] = 0; w6[0] = 0; w6[1] = 0; w6[2] = 0; w6[3] = 0; w7[0] = 0; w7[1] = 0; w7[2] = 0; w7[3] = 0; sha512_hmac_update_128 (&sha512_hmac_ctx2, w0, w1, w2, w3, w4, w5, w6, w7, 4); sha512_hmac_final (&sha512_hmac_ctx2); tmps[gid].dgst[i + 0] = sha512_hmac_ctx2.opad.h[0]; tmps[gid].dgst[i + 1] = sha512_hmac_ctx2.opad.h[1]; tmps[gid].dgst[i + 2] = sha512_hmac_ctx2.opad.h[2]; tmps[gid].dgst[i + 3] = sha512_hmac_ctx2.opad.h[3]; tmps[gid].dgst[i + 4] = sha512_hmac_ctx2.opad.h[4]; tmps[gid].dgst[i + 5] = sha512_hmac_ctx2.opad.h[5]; tmps[gid].dgst[i + 6] = sha512_hmac_ctx2.opad.h[6]; tmps[gid].dgst[i + 7] = sha512_hmac_ctx2.opad.h[7]; tmps[gid].out[i + 0] = tmps[gid].dgst[i + 0]; tmps[gid].out[i + 1] = tmps[gid].dgst[i + 1]; tmps[gid].out[i + 2] = tmps[gid].dgst[i + 2]; tmps[gid].out[i + 3] = tmps[gid].dgst[i + 3]; tmps[gid].out[i + 4] = tmps[gid].dgst[i + 4]; tmps[gid].out[i + 5] = tmps[gid].dgst[i + 5]; tmps[gid].out[i + 6] = tmps[gid].dgst[i + 6]; tmps[gid].out[i + 7] = tmps[gid].dgst[i + 7]; } } __kernel void m06221_loop (__global pw_t *pws, __constant const kernel_rule_t *rules_buf, __global const pw_t *combs_buf, __global const bf_t *bfs_buf, __global tc64_tmp_t *tmps, __global void *hooks, __global const u32 *bitmaps_buf_s1_a, __global const u32 *bitmaps_buf_s1_b, __global const u32 *bitmaps_buf_s1_c, __global const u32 *bitmaps_buf_s1_d, __global const u32 *bitmaps_buf_s2_a, __global const u32 *bitmaps_buf_s2_b, __global const u32 *bitmaps_buf_s2_c, __global const u32 *bitmaps_buf_s2_d, __global plain_t *plains_buf, __global const digest_t *digests_buf, __global u32 *hashes_shown, __global const salt_t *salt_bufs, __global const tc_t *esalt_bufs, __global u32 *d_return_buf, __global u32 *d_scryptV0_buf, __global u32 *d_scryptV1_buf, __global u32 *d_scryptV2_buf, __global u32 *d_scryptV3_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 u64 gid_max) { const u64 gid = get_global_id (0); if ((gid * VECT_SIZE) >= gid_max) return; u64x ipad[8]; u64x opad[8]; ipad[0] = pack64v (tmps, ipad, gid, 0); ipad[1] = pack64v (tmps, ipad, gid, 1); ipad[2] = pack64v (tmps, ipad, gid, 2); ipad[3] = pack64v (tmps, ipad, gid, 3); ipad[4] = pack64v (tmps, ipad, gid, 4); ipad[5] = pack64v (tmps, ipad, gid, 5); ipad[6] = pack64v (tmps, ipad, gid, 6); ipad[7] = pack64v (tmps, ipad, gid, 7); opad[0] = pack64v (tmps, opad, gid, 0); opad[1] = pack64v (tmps, opad, gid, 1); opad[2] = pack64v (tmps, opad, gid, 2); opad[3] = pack64v (tmps, opad, gid, 3); opad[4] = pack64v (tmps, opad, gid, 4); opad[5] = pack64v (tmps, opad, gid, 5); opad[6] = pack64v (tmps, opad, gid, 6); opad[7] = pack64v (tmps, opad, gid, 7); for (u32 i = 0; i < 8; i += 8) { u64x dgst[8]; u64x out[8]; dgst[0] = pack64v (tmps, dgst, gid, i + 0); dgst[1] = pack64v (tmps, dgst, gid, i + 1); dgst[2] = pack64v (tmps, dgst, gid, i + 2); dgst[3] = pack64v (tmps, dgst, gid, i + 3); dgst[4] = pack64v (tmps, dgst, gid, i + 4); dgst[5] = pack64v (tmps, dgst, gid, i + 5); dgst[6] = pack64v (tmps, dgst, gid, i + 6); dgst[7] = pack64v (tmps, dgst, gid, i + 7); out[0] = pack64v (tmps, out, gid, i + 0); out[1] = pack64v (tmps, out, gid, i + 1); out[2] = pack64v (tmps, out, gid, i + 2); out[3] = pack64v (tmps, out, gid, i + 3); out[4] = pack64v (tmps, out, gid, i + 4); out[5] = pack64v (tmps, out, gid, i + 5); out[6] = pack64v (tmps, out, gid, i + 6); out[7] = pack64v (tmps, out, gid, i + 7); for (u32 j = 0; j < loop_cnt; j++) { u32x w0[4]; u32x w1[4]; u32x w2[4]; u32x w3[4]; u32x w4[4]; u32x w5[4]; u32x w6[4]; u32x w7[4]; w0[0] = h32_from_64 (dgst[0]); w0[1] = l32_from_64 (dgst[0]); w0[2] = h32_from_64 (dgst[1]); w0[3] = l32_from_64 (dgst[1]); w1[0] = h32_from_64 (dgst[2]); w1[1] = l32_from_64 (dgst[2]); w1[2] = h32_from_64 (dgst[3]); w1[3] = l32_from_64 (dgst[3]); w2[0] = h32_from_64 (dgst[4]); w2[1] = l32_from_64 (dgst[4]); w2[2] = h32_from_64 (dgst[5]); w2[3] = l32_from_64 (dgst[5]); w3[0] = h32_from_64 (dgst[6]); w3[1] = l32_from_64 (dgst[6]); w3[2] = h32_from_64 (dgst[7]); w3[3] = l32_from_64 (dgst[7]); w4[0] = 0x80000000; w4[1] = 0; w4[2] = 0; w4[3] = 0; w5[0] = 0; w5[1] = 0; w5[2] = 0; w5[3] = 0; w6[0] = 0; w6[1] = 0; w6[2] = 0; w6[3] = 0; w7[0] = 0; w7[1] = 0; w7[2] = 0; w7[3] = (128 + 64) * 8; hmac_sha512_run_V (w0, w1, w2, w3, w4, w5, w6, w7, ipad, opad, dgst); out[0] ^= dgst[0]; out[1] ^= dgst[1]; out[2] ^= dgst[2]; out[3] ^= dgst[3]; out[4] ^= dgst[4]; out[5] ^= dgst[5]; out[6] ^= dgst[6]; out[7] ^= dgst[7]; } unpack64v (tmps, dgst, gid, i + 0, dgst[0]); unpack64v (tmps, dgst, gid, i + 1, dgst[1]); unpack64v (tmps, dgst, gid, i + 2, dgst[2]); unpack64v (tmps, dgst, gid, i + 3, dgst[3]); unpack64v (tmps, dgst, gid, i + 4, dgst[4]); unpack64v (tmps, dgst, gid, i + 5, dgst[5]); unpack64v (tmps, dgst, gid, i + 6, dgst[6]); unpack64v (tmps, dgst, gid, i + 7, dgst[7]); unpack64v (tmps, out, gid, i + 0, out[0]); unpack64v (tmps, out, gid, i + 1, out[1]); unpack64v (tmps, out, gid, i + 2, out[2]); unpack64v (tmps, out, gid, i + 3, out[3]); unpack64v (tmps, out, gid, i + 4, out[4]); unpack64v (tmps, out, gid, i + 5, out[5]); unpack64v (tmps, out, gid, i + 6, out[6]); unpack64v (tmps, out, gid, i + 7, out[7]); } } __kernel void m06221_comp (__global pw_t *pws, __constant const kernel_rule_t *rules_buf, __global const pw_t *combs_buf, __global const bf_t *bfs_buf, __global tc64_tmp_t *tmps, __global void *hooks, __global const u32 *bitmaps_buf_s1_a, __global const u32 *bitmaps_buf_s1_b, __global const u32 *bitmaps_buf_s1_c, __global const u32 *bitmaps_buf_s1_d, __global const u32 *bitmaps_buf_s2_a, __global const u32 *bitmaps_buf_s2_b, __global const u32 *bitmaps_buf_s2_c, __global const u32 *bitmaps_buf_s2_d, __global plain_t *plains_buf, __global const digest_t *digests_buf, __global u32 *hashes_shown, __global const salt_t *salt_bufs, __global const tc_t *esalt_bufs, __global u32 *d_return_buf, __global u32 *d_scryptV0_buf, __global u32 *d_scryptV1_buf, __global u32 *d_scryptV2_buf, __global u32 *d_scryptV3_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 u64 gid_max) { const u64 gid = get_global_id (0); const u64 lid = get_local_id (0); const u64 lsz = get_local_size (0); /** * aes shared */ #ifdef REAL_SHM __local u32 s_td0[256]; __local u32 s_td1[256]; __local u32 s_td2[256]; __local u32 s_td3[256]; __local u32 s_td4[256]; __local u32 s_te0[256]; __local u32 s_te1[256]; __local u32 s_te2[256]; __local u32 s_te3[256]; __local u32 s_te4[256]; for (u32 i = lid; i < 256; i += lsz) { s_td0[i] = td0[i]; s_td1[i] = td1[i]; s_td2[i] = td2[i]; s_td3[i] = td3[i]; s_td4[i] = td4[i]; s_te0[i] = te0[i]; s_te1[i] = te1[i]; s_te2[i] = te2[i]; s_te3[i] = te3[i]; s_te4[i] = te4[i]; } barrier (CLK_LOCAL_MEM_FENCE); #else __constant u32a *s_td0 = td0; __constant u32a *s_td1 = td1; __constant u32a *s_td2 = td2; __constant u32a *s_td3 = td3; __constant u32a *s_td4 = td4; __constant u32a *s_te0 = te0; __constant u32a *s_te1 = te1; __constant u32a *s_te2 = te2; __constant u32a *s_te3 = te3; __constant u32a *s_te4 = te4; #endif if (gid >= gid_max) return; u32 ukey1[8]; ukey1[0] = swap32_S (h32_from_64_S (tmps[gid].out[0])); ukey1[1] = swap32_S (l32_from_64_S (tmps[gid].out[0])); ukey1[2] = swap32_S (h32_from_64_S (tmps[gid].out[1])); ukey1[3] = swap32_S (l32_from_64_S (tmps[gid].out[1])); ukey1[4] = swap32_S (h32_from_64_S (tmps[gid].out[2])); ukey1[5] = swap32_S (l32_from_64_S (tmps[gid].out[2])); ukey1[6] = swap32_S (h32_from_64_S (tmps[gid].out[3])); ukey1[7] = swap32_S (l32_from_64_S (tmps[gid].out[3])); u32 ukey2[8]; ukey2[0] = swap32_S (h32_from_64_S (tmps[gid].out[4])); ukey2[1] = swap32_S (l32_from_64_S (tmps[gid].out[4])); ukey2[2] = swap32_S (h32_from_64_S (tmps[gid].out[5])); ukey2[3] = swap32_S (l32_from_64_S (tmps[gid].out[5])); ukey2[4] = swap32_S (h32_from_64_S (tmps[gid].out[6])); ukey2[5] = swap32_S (l32_from_64_S (tmps[gid].out[6])); ukey2[6] = swap32_S (h32_from_64_S (tmps[gid].out[7])); ukey2[7] = swap32_S (l32_from_64_S (tmps[gid].out[7])); if (verify_header_aes (esalt_bufs, ukey1, ukey2, s_te0, s_te1, s_te2, s_te3, s_te4, s_td0, s_td1, s_td2, s_td3, s_td4) == 1) { if (atomic_inc (&hashes_shown[0]) == 0) { mark_hash (plains_buf, d_return_buf, salt_pos, digests_cnt, 0, 0, gid, 0); } } if (verify_header_serpent (esalt_bufs, ukey1, ukey2) == 1) { if (atomic_inc (&hashes_shown[0]) == 0) { mark_hash (plains_buf, d_return_buf, salt_pos, digests_cnt, 0, 0, gid, 0); } } if (verify_header_twofish (esalt_bufs, ukey1, ukey2) == 1) { if (atomic_inc (&hashes_shown[0]) == 0) { mark_hash (plains_buf, d_return_buf, salt_pos, digests_cnt, 0, 0, gid, 0); } } }