1
0
mirror of https://github.com/hashcat/hashcat.git synced 2024-12-30 18:31:00 +00:00
hashcat/OpenCL/m07300_a0-optimized.cl

372 lines
10 KiB
Common Lisp
Raw Normal View History

2015-12-04 14:47:52 +00:00
/**
* Author......: See docs/credits.txt
2015-12-04 14:47:52 +00:00
* License.....: MIT
*/
2016-02-06 09:47:19 +00:00
#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"
2017-08-11 09:25:47 +00:00
#include "inc_rp_optimized.h"
#include "inc_rp_optimized.cl"
#include "inc_simd.cl"
#include "inc_hash_sha1.cl"
2015-12-04 14:47:52 +00:00
DECLSPEC void hmac_sha1_pad (u32x *w0, u32x *w1, u32x *w2, u32x *w3, u32x *ipad, u32x *opad)
2015-12-04 14:47:52 +00:00
{
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_vector (w0, w1, w2, w3, ipad);
2015-12-04 14:47:52 +00:00
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_vector (w0, w1, w2, w3, opad);
2015-12-04 14:47:52 +00:00
}
DECLSPEC void hmac_sha1_run (u32x *w0, u32x *w1, u32x *w2, u32x *w3, u32x *ipad, u32x *opad, u32x *digest)
2015-12-04 14:47:52 +00:00
{
digest[0] = ipad[0];
digest[1] = ipad[1];
digest[2] = ipad[2];
digest[3] = ipad[3];
digest[4] = ipad[4];
sha1_transform_vector (w0, w1, w2, w3, digest);
2015-12-04 14:47:52 +00:00
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_vector (w0, w1, w2, w3, digest);
2015-12-04 14:47:52 +00:00
}
__kernel void m07300_m04 (KERN_ATTR_RULES_ESALT (rakp_t))
2015-12-04 14:47:52 +00:00
{
/**
* modifier
*/
const u64 lid = get_local_id (0);
2015-12-04 14:47:52 +00:00
/**
* base
*/
const u64 gid = get_global_id (0);
2015-12-04 14:47:52 +00:00
if (gid >= gid_max) return;
u32 pw_buf0[4];
u32 pw_buf1[4];
2015-12-04 14:47:52 +00:00
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];
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];
2015-12-04 14:47:52 +00:00
const u32 pw_len = pws[gid].pw_len & 63;
2015-12-04 14:47:52 +00:00
/**
* salt
*/
const u32 esalt_len = esalt_bufs[digests_offset].salt_len;
2015-12-04 14:47:52 +00:00
/**
* loop
*/
for (u32 il_pos = 0; il_pos < il_cnt; il_pos += VECT_SIZE)
2015-12-04 14:47:52 +00:00
{
2016-02-06 09:47:19 +00:00
u32x w0[4] = { 0 };
u32x w1[4] = { 0 };
u32x w2[4] = { 0 };
u32x w3[4] = { 0 };
2015-12-04 14:47:52 +00:00
apply_rules_vect (pw_buf0, pw_buf1, pw_len, rules_buf, il_pos, w0, w1);
2015-12-04 14:47:52 +00:00
w0[0] = swap32 (w0[0]);
w0[1] = swap32 (w0[1]);
w0[2] = swap32 (w0[2]);
w0[3] = swap32 (w0[3]);
w1[0] = swap32 (w1[0]);
w1[1] = swap32 (w1[1]);
w1[2] = swap32 (w1[2]);
w1[3] = swap32 (w1[3]);
2015-12-04 14:47:52 +00:00
/**
* RAKP
2015-12-04 14:47:52 +00:00
*/
2016-02-06 09:47:19 +00:00
u32x ipad[5];
u32x opad[5];
2015-12-04 14:47:52 +00:00
hmac_sha1_pad (w0, w1, w2, w3, ipad, opad);
2015-12-04 14:47:52 +00:00
int esalt_size = esalt_len;
int esalt_left;
int esalt_off;
for (esalt_left = esalt_size, esalt_off = 0; esalt_left >= 56; esalt_left -= 64, esalt_off += 16)
{
w0[0] = esalt_bufs[digests_offset].salt_buf[esalt_off + 0];
w0[1] = esalt_bufs[digests_offset].salt_buf[esalt_off + 1];
w0[2] = esalt_bufs[digests_offset].salt_buf[esalt_off + 2];
w0[3] = esalt_bufs[digests_offset].salt_buf[esalt_off + 3];
w1[0] = esalt_bufs[digests_offset].salt_buf[esalt_off + 4];
w1[1] = esalt_bufs[digests_offset].salt_buf[esalt_off + 5];
w1[2] = esalt_bufs[digests_offset].salt_buf[esalt_off + 6];
w1[3] = esalt_bufs[digests_offset].salt_buf[esalt_off + 7];
w2[0] = esalt_bufs[digests_offset].salt_buf[esalt_off + 8];
w2[1] = esalt_bufs[digests_offset].salt_buf[esalt_off + 9];
w2[2] = esalt_bufs[digests_offset].salt_buf[esalt_off + 10];
w2[3] = esalt_bufs[digests_offset].salt_buf[esalt_off + 11];
w3[0] = esalt_bufs[digests_offset].salt_buf[esalt_off + 12];
w3[1] = esalt_bufs[digests_offset].salt_buf[esalt_off + 13];
w3[2] = esalt_bufs[digests_offset].salt_buf[esalt_off + 14];
w3[3] = esalt_bufs[digests_offset].salt_buf[esalt_off + 15];
sha1_transform_vector (w0, w1, w2, w3, ipad);
2015-12-04 14:47:52 +00:00
}
w0[0] = esalt_bufs[digests_offset].salt_buf[esalt_off + 0];
w0[1] = esalt_bufs[digests_offset].salt_buf[esalt_off + 1];
w0[2] = esalt_bufs[digests_offset].salt_buf[esalt_off + 2];
w0[3] = esalt_bufs[digests_offset].salt_buf[esalt_off + 3];
w1[0] = esalt_bufs[digests_offset].salt_buf[esalt_off + 4];
w1[1] = esalt_bufs[digests_offset].salt_buf[esalt_off + 5];
w1[2] = esalt_bufs[digests_offset].salt_buf[esalt_off + 6];
w1[3] = esalt_bufs[digests_offset].salt_buf[esalt_off + 7];
w2[0] = esalt_bufs[digests_offset].salt_buf[esalt_off + 8];
w2[1] = esalt_bufs[digests_offset].salt_buf[esalt_off + 9];
w2[2] = esalt_bufs[digests_offset].salt_buf[esalt_off + 10];
w2[3] = esalt_bufs[digests_offset].salt_buf[esalt_off + 11];
w3[0] = esalt_bufs[digests_offset].salt_buf[esalt_off + 12];
w3[1] = esalt_bufs[digests_offset].salt_buf[esalt_off + 13];
w3[2] = 0;
w3[3] = (64 + esalt_size) * 8;
2015-12-04 14:47:52 +00:00
2016-02-06 09:47:19 +00:00
u32x digest[5];
2015-12-04 14:47:52 +00:00
hmac_sha1_run (w0, w1, w2, w3, ipad, opad, digest);
2015-12-04 14:47:52 +00:00
2016-02-06 09:47:19 +00:00
COMPARE_M_SIMD (digest[3], digest[4], digest[2], digest[1]);
2015-12-04 14:47:52 +00:00
}
}
__kernel void m07300_m08 (KERN_ATTR_RULES_ESALT (rakp_t))
2015-12-04 14:47:52 +00:00
{
}
__kernel void m07300_m16 (KERN_ATTR_RULES_ESALT (rakp_t))
2015-12-04 14:47:52 +00:00
{
}
__kernel void m07300_s04 (KERN_ATTR_RULES_ESALT (rakp_t))
2015-12-04 14:47:52 +00:00
{
/**
* modifier
*/
const u64 lid = get_local_id (0);
2015-12-04 14:47:52 +00:00
/**
* base
*/
const u64 gid = get_global_id (0);
2015-12-04 14:47:52 +00:00
if (gid >= gid_max) return;
u32 pw_buf0[4];
u32 pw_buf1[4];
2015-12-04 14:47:52 +00:00
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];
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];
2015-12-04 14:47:52 +00:00
const u32 pw_len = pws[gid].pw_len & 63;
2015-12-04 14:47:52 +00:00
/**
* salt
*/
const u32 esalt_len = esalt_bufs[digests_offset].salt_len;
2015-12-04 14:47:52 +00:00
/**
* 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 < il_cnt; il_pos += VECT_SIZE)
2015-12-04 14:47:52 +00:00
{
2016-02-06 09:47:19 +00:00
u32x w0[4] = { 0 };
u32x w1[4] = { 0 };
u32x w2[4] = { 0 };
u32x w3[4] = { 0 };
2015-12-04 14:47:52 +00:00
apply_rules_vect (pw_buf0, pw_buf1, pw_len, rules_buf, il_pos, w0, w1);
2015-12-04 14:47:52 +00:00
w0[0] = swap32 (w0[0]);
w0[1] = swap32 (w0[1]);
w0[2] = swap32 (w0[2]);
w0[3] = swap32 (w0[3]);
w1[0] = swap32 (w1[0]);
w1[1] = swap32 (w1[1]);
w1[2] = swap32 (w1[2]);
w1[3] = swap32 (w1[3]);
2015-12-04 14:47:52 +00:00
/**
* RAKP
2015-12-04 14:47:52 +00:00
*/
2016-02-06 09:47:19 +00:00
u32x ipad[5];
u32x opad[5];
2015-12-04 14:47:52 +00:00
hmac_sha1_pad (w0, w1, w2, w3, ipad, opad);
2015-12-04 14:47:52 +00:00
int esalt_size = esalt_len;
int esalt_left;
int esalt_off;
for (esalt_left = esalt_size, esalt_off = 0; esalt_left >= 56; esalt_left -= 64, esalt_off += 16)
{
w0[0] = esalt_bufs[digests_offset].salt_buf[esalt_off + 0];
w0[1] = esalt_bufs[digests_offset].salt_buf[esalt_off + 1];
w0[2] = esalt_bufs[digests_offset].salt_buf[esalt_off + 2];
w0[3] = esalt_bufs[digests_offset].salt_buf[esalt_off + 3];
w1[0] = esalt_bufs[digests_offset].salt_buf[esalt_off + 4];
w1[1] = esalt_bufs[digests_offset].salt_buf[esalt_off + 5];
w1[2] = esalt_bufs[digests_offset].salt_buf[esalt_off + 6];
w1[3] = esalt_bufs[digests_offset].salt_buf[esalt_off + 7];
w2[0] = esalt_bufs[digests_offset].salt_buf[esalt_off + 8];
w2[1] = esalt_bufs[digests_offset].salt_buf[esalt_off + 9];
w2[2] = esalt_bufs[digests_offset].salt_buf[esalt_off + 10];
w2[3] = esalt_bufs[digests_offset].salt_buf[esalt_off + 11];
w3[0] = esalt_bufs[digests_offset].salt_buf[esalt_off + 12];
w3[1] = esalt_bufs[digests_offset].salt_buf[esalt_off + 13];
w3[2] = esalt_bufs[digests_offset].salt_buf[esalt_off + 14];
w3[3] = esalt_bufs[digests_offset].salt_buf[esalt_off + 15];
sha1_transform_vector (w0, w1, w2, w3, ipad);
2015-12-04 14:47:52 +00:00
}
w0[0] = esalt_bufs[digests_offset].salt_buf[esalt_off + 0];
w0[1] = esalt_bufs[digests_offset].salt_buf[esalt_off + 1];
w0[2] = esalt_bufs[digests_offset].salt_buf[esalt_off + 2];
w0[3] = esalt_bufs[digests_offset].salt_buf[esalt_off + 3];
w1[0] = esalt_bufs[digests_offset].salt_buf[esalt_off + 4];
w1[1] = esalt_bufs[digests_offset].salt_buf[esalt_off + 5];
w1[2] = esalt_bufs[digests_offset].salt_buf[esalt_off + 6];
w1[3] = esalt_bufs[digests_offset].salt_buf[esalt_off + 7];
w2[0] = esalt_bufs[digests_offset].salt_buf[esalt_off + 8];
w2[1] = esalt_bufs[digests_offset].salt_buf[esalt_off + 9];
w2[2] = esalt_bufs[digests_offset].salt_buf[esalt_off + 10];
w2[3] = esalt_bufs[digests_offset].salt_buf[esalt_off + 11];
w3[0] = esalt_bufs[digests_offset].salt_buf[esalt_off + 12];
w3[1] = esalt_bufs[digests_offset].salt_buf[esalt_off + 13];
w3[2] = 0;
w3[3] = (64 + esalt_size) * 8;
2015-12-04 14:47:52 +00:00
2016-02-06 09:47:19 +00:00
u32x digest[5];
2015-12-04 14:47:52 +00:00
hmac_sha1_run (w0, w1, w2, w3, ipad, opad, digest);
2015-12-04 14:47:52 +00:00
2016-02-06 09:47:19 +00:00
COMPARE_S_SIMD (digest[3], digest[4], digest[2], digest[1]);
2015-12-04 14:47:52 +00:00
}
}
__kernel void m07300_s08 (KERN_ATTR_RULES_ESALT (rakp_t))
2015-12-04 14:47:52 +00:00
{
}
__kernel void m07300_s16 (KERN_ATTR_RULES_ESALT (rakp_t))
2015-12-04 14:47:52 +00:00
{
}