1
mirror of https://github.com/hashcat/hashcat synced 2024-11-20 23:27:31 +01:00
hashcat/OpenCL/m13100_a1.cl

493 lines
12 KiB
Common Lisp

/**
* Author......: See docs/credits.txt
* License.....: MIT
*/
//shared mem too small
//#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_hash_md4.cl"
#include "inc_hash_md5.cl"
typedef struct
{
u8 S[256];
u32 wtf_its_faster;
} RC4_KEY;
void swap (__local RC4_KEY *rc4_key, const u8 i, const u8 j)
{
u8 tmp;
tmp = rc4_key->S[i];
rc4_key->S[i] = rc4_key->S[j];
rc4_key->S[j] = tmp;
}
void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
{
u32 v = 0x03020100;
u32 a = 0x04040404;
__local u32 *ptr = (__local u32 *) rc4_key->S;
#ifdef _unroll
#pragma unroll
#endif
for (u32 i = 0; i < 64; i++)
{
*ptr++ = v; v += a;
}
u32 j = 0;
for (u32 i = 0; i < 16; i++)
{
u32 idx = i * 16;
u32 v;
v = data[0];
j += rc4_key->S[idx] + (v >> 0); swap (rc4_key, idx, j); idx++;
j += rc4_key->S[idx] + (v >> 8); swap (rc4_key, idx, j); idx++;
j += rc4_key->S[idx] + (v >> 16); swap (rc4_key, idx, j); idx++;
j += rc4_key->S[idx] + (v >> 24); swap (rc4_key, idx, j); idx++;
v = data[1];
j += rc4_key->S[idx] + (v >> 0); swap (rc4_key, idx, j); idx++;
j += rc4_key->S[idx] + (v >> 8); swap (rc4_key, idx, j); idx++;
j += rc4_key->S[idx] + (v >> 16); swap (rc4_key, idx, j); idx++;
j += rc4_key->S[idx] + (v >> 24); swap (rc4_key, idx, j); idx++;
v = data[2];
j += rc4_key->S[idx] + (v >> 0); swap (rc4_key, idx, j); idx++;
j += rc4_key->S[idx] + (v >> 8); swap (rc4_key, idx, j); idx++;
j += rc4_key->S[idx] + (v >> 16); swap (rc4_key, idx, j); idx++;
j += rc4_key->S[idx] + (v >> 24); swap (rc4_key, idx, j); idx++;
v = data[3];
j += rc4_key->S[idx] + (v >> 0); swap (rc4_key, idx, j); idx++;
j += rc4_key->S[idx] + (v >> 8); swap (rc4_key, idx, j); idx++;
j += rc4_key->S[idx] + (v >> 16); swap (rc4_key, idx, j); idx++;
j += rc4_key->S[idx] + (v >> 24); swap (rc4_key, idx, j); idx++;
}
}
u8 rc4_next_16 (__local RC4_KEY *rc4_key, u8 i, u8 j, __global const u32 in[4], u32 out[4])
{
#ifdef _unroll
#pragma unroll
#endif
for (u32 k = 0; k < 4; k++)
{
u32 xor4 = 0;
u8 idx;
i += 1;
j += rc4_key->S[i];
swap (rc4_key, i, j);
idx = rc4_key->S[i] + rc4_key->S[j];
xor4 |= rc4_key->S[idx] << 0;
i += 1;
j += rc4_key->S[i];
swap (rc4_key, i, j);
idx = rc4_key->S[i] + rc4_key->S[j];
xor4 |= rc4_key->S[idx] << 8;
i += 1;
j += rc4_key->S[i];
swap (rc4_key, i, j);
idx = rc4_key->S[i] + rc4_key->S[j];
xor4 |= rc4_key->S[idx] << 16;
i += 1;
j += rc4_key->S[i];
swap (rc4_key, i, j);
idx = rc4_key->S[i] + rc4_key->S[j];
xor4 |= rc4_key->S[idx] << 24;
out[k] = in[k] ^ xor4;
}
return j;
}
int decrypt_and_check (__local RC4_KEY *rc4_key, u32 data[4], __global const u32 *edata2, const u32 edata2_len, const u32 K2[4], const u32 checksum[4])
{
rc4_init_16 (rc4_key, data);
u32 out0[4];
u32 out1[4];
u8 i = 0;
u8 j = 0;
/*
8 first bytes are nonce, then ASN1 structs (DER encoding: type-length-data)
if length >= 128 bytes:
length is on 2 bytes and type is \x63\x82 (encode_krb5_enc_tkt_part) and data is an ASN1 sequence \x30\x82
else:
length is on 1 byte and type is \x63\x81 and data is an ASN1 sequence \x30\x81
next headers follow the same ASN1 "type-length-data" scheme
*/
j = rc4_next_16 (rc4_key, i, j, edata2 + 0, out0); i += 16;
if (((out0[2] & 0xff00ffff) != 0x30008163) && ((out0[2] & 0x0000ffff) != 0x00008263)) return 0;
j = rc4_next_16 (rc4_key, i, j, edata2 + 4, out1); i += 16;
if (((out1[0] & 0x00ffffff) != 0x00000503) && (out1[0] != 0x050307A0)) return 0;
rc4_init_16 (rc4_key, data);
i = 0;
j = 0;
// init hmac
u32 w0[4];
u32 w1[4];
u32 w2[4];
u32 w3[4];
w0[0] = K2[0];
w0[1] = K2[1];
w0[2] = K2[2];
w0[3] = K2[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;
md5_hmac_ctx_t ctx;
md5_hmac_init_64 (&ctx, w0, w1, w2, w3);
int edata2_left;
for (edata2_left = edata2_len; edata2_left >= 64; edata2_left -= 64)
{
j = rc4_next_16 (rc4_key, i, j, edata2, w0); i += 16; edata2 += 4;
j = rc4_next_16 (rc4_key, i, j, edata2, w1); i += 16; edata2 += 4;
j = rc4_next_16 (rc4_key, i, j, edata2, w2); i += 16; edata2 += 4;
j = rc4_next_16 (rc4_key, i, j, edata2, w3); i += 16; edata2 += 4;
md5_hmac_update_64 (&ctx, w0, w1, w2, w3, 64);
}
w0[0] = 0;
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;
if (edata2_left < 16)
{
j = rc4_next_16 (rc4_key, i, j, edata2, w0); i += 16; edata2 += 4;
truncate_block_4x4_le (w0, edata2_left & 0xf);
}
else if (edata2_left < 32)
{
j = rc4_next_16 (rc4_key, i, j, edata2, w0); i += 16; edata2 += 4;
j = rc4_next_16 (rc4_key, i, j, edata2, w1); i += 16; edata2 += 4;
truncate_block_4x4_le (w1, edata2_left & 0xf);
}
else if (edata2_left < 48)
{
j = rc4_next_16 (rc4_key, i, j, edata2, w0); i += 16; edata2 += 4;
j = rc4_next_16 (rc4_key, i, j, edata2, w1); i += 16; edata2 += 4;
j = rc4_next_16 (rc4_key, i, j, edata2, w2); i += 16; edata2 += 4;
truncate_block_4x4_le (w2, edata2_left & 0xf);
}
else
{
j = rc4_next_16 (rc4_key, i, j, edata2, w0); i += 16; edata2 += 4;
j = rc4_next_16 (rc4_key, i, j, edata2, w1); i += 16; edata2 += 4;
j = rc4_next_16 (rc4_key, i, j, edata2, w2); i += 16; edata2 += 4;
j = rc4_next_16 (rc4_key, i, j, edata2, w3); i += 16; edata2 += 4;
truncate_block_4x4_le (w3, edata2_left & 0xf);
}
md5_hmac_update_64 (&ctx, w0, w1, w2, w3, edata2_left);
md5_hmac_final (&ctx);
if (checksum[0] != ctx.opad.h[0]) return 0;
if (checksum[1] != ctx.opad.h[1]) return 0;
if (checksum[2] != ctx.opad.h[2]) return 0;
if (checksum[3] != ctx.opad.h[3]) return 0;
return 1;
}
void kerb_prepare (const u32 K[4], const u32 checksum[4], u32 digest[4], u32 K2[4])
{
// K1=MD5_HMAC(K,1); with 1 encoded as little indian on 4 bytes (01000000 in hexa);
u32 w0[4];
u32 w1[4];
u32 w2[4];
u32 w3[4];
w0[0] = K[0];
w0[1] = K[1];
w0[2] = K[2];
w0[3] = K[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;
md5_hmac_ctx_t ctx1;
md5_hmac_init_64 (&ctx1, w0, w1, w2, w3);
w0[0] = 2;
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;
md5_hmac_update_64 (&ctx1, w0, w1, w2, w3, 4);
md5_hmac_final (&ctx1);
w0[0] = ctx1.opad.h[0];
w0[1] = ctx1.opad.h[1];
w0[2] = ctx1.opad.h[2];
w0[3] = ctx1.opad.h[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;
md5_hmac_ctx_t ctx;
md5_hmac_init_64 (&ctx, w0, w1, w2, w3);
w0[0] = checksum[0];
w0[1] = checksum[1];
w0[2] = checksum[2];
w0[3] = checksum[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;
md5_hmac_update_64 (&ctx, w0, w1, w2, w3, 16);
md5_hmac_final (&ctx);
digest[0] = ctx.opad.h[0];
digest[1] = ctx.opad.h[1];
digest[2] = ctx.opad.h[2];
digest[3] = ctx.opad.h[3];
K2[0] = ctx1.opad.h[0];
K2[1] = ctx1.opad.h[1];
K2[2] = ctx1.opad.h[2];
K2[3] = ctx1.opad.h[3];
}
__kernel void m13100_mxx (__global pw_t *pws, __constant const kernel_rule_t *rules_buf, __global const pw_t *combs_buf, __global const bf_t *bfs_buf, __global void *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 krb5tgs_t *krb5tgs_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)
{
/**
* modifier
*/
const u64 lid = get_local_id (0);
const u64 gid = get_global_id (0);
if (gid >= gid_max) return;
/**
* base
*/
__local RC4_KEY rc4_keys[64];
u32 checksum[4];
checksum[0] = krb5tgs_bufs[digests_offset].checksum[0];
checksum[1] = krb5tgs_bufs[digests_offset].checksum[1];
checksum[2] = krb5tgs_bufs[digests_offset].checksum[2];
checksum[3] = krb5tgs_bufs[digests_offset].checksum[3];
md4_ctx_t ctx0;
md4_init (&ctx0);
md4_update_global_utf16le (&ctx0, pws[gid].i, pws[gid].pw_len);
/**
* loop
*/
for (u32 il_pos = 0; il_pos < il_cnt; il_pos++)
{
md4_ctx_t ctx = ctx0;
md4_update_global_utf16le (&ctx, combs_buf[il_pos].i, combs_buf[il_pos].pw_len);
md4_final (&ctx);
u32 digest[4];
u32 K2[4];
kerb_prepare (ctx.h, checksum, digest, K2);
if (decrypt_and_check (&rc4_keys[lid], digest, krb5tgs_bufs[digests_offset].edata2, krb5tgs_bufs[digests_offset].edata2_len, K2, checksum) == 1)
{
if (atomic_inc (&hashes_shown[digests_offset]) == 0)
{
mark_hash (plains_buf, d_return_buf, salt_pos, digests_cnt, 0, digests_offset + 0, gid, il_pos);
}
}
}
}
__kernel void m13100_sxx (__global pw_t *pws, __constant const kernel_rule_t *rules_buf, __global const pw_t *combs_buf, __global const bf_t *bfs_buf, __global void *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 krb5tgs_t *krb5tgs_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)
{
/**
* modifier
*/
const u64 lid = get_local_id (0);
const u64 gid = get_global_id (0);
if (gid >= gid_max) return;
/**
* base
*/
__local RC4_KEY rc4_keys[64];
u32 checksum[4];
checksum[0] = krb5tgs_bufs[digests_offset].checksum[0];
checksum[1] = krb5tgs_bufs[digests_offset].checksum[1];
checksum[2] = krb5tgs_bufs[digests_offset].checksum[2];
checksum[3] = krb5tgs_bufs[digests_offset].checksum[3];
md4_ctx_t ctx0;
md4_init (&ctx0);
md4_update_global_utf16le (&ctx0, pws[gid].i, pws[gid].pw_len);
/**
* loop
*/
for (u32 il_pos = 0; il_pos < il_cnt; il_pos++)
{
md4_ctx_t ctx = ctx0;
md4_update_global_utf16le (&ctx, combs_buf[il_pos].i, combs_buf[il_pos].pw_len);
md4_final (&ctx);
u32 digest[4];
u32 K2[4];
kerb_prepare (ctx.h, checksum, digest, K2);
if (decrypt_and_check (&rc4_keys[lid], digest, krb5tgs_bufs[digests_offset].edata2, krb5tgs_bufs[digests_offset].edata2_len, K2, checksum) == 1)
{
if (atomic_inc (&hashes_shown[digests_offset]) == 0)
{
mark_hash (plains_buf, d_return_buf, salt_pos, digests_cnt, 0, digests_offset + 0, gid, il_pos);
}
}
}
}