|
|
|
/**
|
|
|
|
* Author......: See docs/credits.txt
|
|
|
|
* License.....: MIT
|
|
|
|
*/
|
|
|
|
|
|
|
|
#define NEW_SIMD_CODE
|
|
|
|
|
|
|
|
#ifdef KERNEL_STATIC
|
|
|
|
#include "inc_vendor.h"
|
|
|
|
#include "inc_types.h"
|
|
|
|
#include "inc_platform.cl"
|
|
|
|
#include "inc_common.cl"
|
|
|
|
#include "inc_simd.cl"
|
|
|
|
#include "inc_hash_sha1.cl"
|
|
|
|
#include "inc_cipher_des.cl"
|
|
|
|
#endif
|
|
|
|
|
|
|
|
typedef struct keychain_tmp
|
|
|
|
{
|
|
|
|
u32 ipad[5];
|
|
|
|
u32 opad[5];
|
|
|
|
|
|
|
|
u32 dgst[10];
|
|
|
|
u32 out [10];
|
|
|
|
|
|
|
|
} keychain_tmp_t;
|
|
|
|
|
|
|
|
typedef struct keychain
|
|
|
|
{
|
|
|
|
u32 data[12];
|
|
|
|
u32 iv[2];
|
|
|
|
|
|
|
|
} keychain_t;
|
|
|
|
|
|
|
|
DECLSPEC void hmac_sha1_run_V (u32x *w0, u32x *w1, u32x *w2, u32x *w3, u32x *ipad, u32x *opad, u32x *digest)
|
|
|
|
{
|
|
|
|
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);
|
|
|
|
|
|
|
|
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);
|
|
|
|
}
|
|
|
|
|
|
|
|
KERNEL_FQ void m23100_init (KERN_ATTR_TMPS_ESALT (keychain_tmp_t, keychain_t))
|
|
|
|
{
|
|
|
|
/**
|
|
|
|
* base
|
|
|
|
*/
|
|
|
|
|
|
|
|
const u64 gid = get_global_id (0);
|
|
|
|
|
|
|
|
if (gid >= GID_CNT) return;
|
|
|
|
|
|
|
|
sha1_hmac_ctx_t sha1_hmac_ctx;
|
|
|
|
|
|
|
|
sha1_hmac_init_global_swap (&sha1_hmac_ctx, pws[gid].i, pws[gid].pw_len);
|
|
|
|
|
|
|
|
tmps[gid].ipad[0] = sha1_hmac_ctx.ipad.h[0];
|
|
|
|
tmps[gid].ipad[1] = sha1_hmac_ctx.ipad.h[1];
|
|
|
|
tmps[gid].ipad[2] = sha1_hmac_ctx.ipad.h[2];
|
|
|
|
tmps[gid].ipad[3] = sha1_hmac_ctx.ipad.h[3];
|
|
|
|
tmps[gid].ipad[4] = sha1_hmac_ctx.ipad.h[4];
|
|
|
|
|
|
|
|
tmps[gid].opad[0] = sha1_hmac_ctx.opad.h[0];
|
|
|
|
tmps[gid].opad[1] = sha1_hmac_ctx.opad.h[1];
|
|
|
|
tmps[gid].opad[2] = sha1_hmac_ctx.opad.h[2];
|
|
|
|
tmps[gid].opad[3] = sha1_hmac_ctx.opad.h[3];
|
|
|
|
tmps[gid].opad[4] = sha1_hmac_ctx.opad.h[4];
|
|
|
|
|
|
|
|
sha1_hmac_update_global_swap (&sha1_hmac_ctx, salt_bufs[SALT_POS_HOST].salt_buf, 20);
|
|
|
|
|
|
|
|
for (u32 i = 0, j = 1; i < 8; i += 5, j += 1)
|
|
|
|
{
|
|
|
|
sha1_hmac_ctx_t sha1_hmac_ctx2 = sha1_hmac_ctx;
|
|
|
|
|
|
|
|
u32 w0[4];
|
|
|
|
u32 w1[4];
|
|
|
|
u32 w2[4];
|
|
|
|
u32 w3[4];
|
|
|
|
|
|
|
|
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;
|
|
|
|
|
|
|
|
sha1_hmac_update_64 (&sha1_hmac_ctx2, w0, w1, w2, w3, 4);
|
|
|
|
|
|
|
|
sha1_hmac_final (&sha1_hmac_ctx2);
|
|
|
|
|
|
|
|
tmps[gid].dgst[i + 0] = sha1_hmac_ctx2.opad.h[0];
|
|
|
|
tmps[gid].dgst[i + 1] = sha1_hmac_ctx2.opad.h[1];
|
|
|
|
tmps[gid].dgst[i + 2] = sha1_hmac_ctx2.opad.h[2];
|
|
|
|
tmps[gid].dgst[i + 3] = sha1_hmac_ctx2.opad.h[3];
|
|
|
|
tmps[gid].dgst[i + 4] = sha1_hmac_ctx2.opad.h[4];
|
|
|
|
|
|
|
|
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];
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
KERNEL_FQ void m23100_loop (KERN_ATTR_TMPS_ESALT (keychain_tmp_t, keychain_t))
|
|
|
|
{
|
|
|
|
const u64 gid = get_global_id (0);
|
|
|
|
|
|
|
|
if ((gid * VECT_SIZE) >= GID_CNT) return;
|
|
|
|
|
|
|
|
u32x ipad[5];
|
|
|
|
u32x opad[5];
|
|
|
|
|
|
|
|
ipad[0] = packv (tmps, ipad, gid, 0);
|
|
|
|
ipad[1] = packv (tmps, ipad, gid, 1);
|
|
|
|
ipad[2] = packv (tmps, ipad, gid, 2);
|
|
|
|
ipad[3] = packv (tmps, ipad, gid, 3);
|
|
|
|
ipad[4] = packv (tmps, ipad, gid, 4);
|
|
|
|
|
|
|
|
opad[0] = packv (tmps, opad, gid, 0);
|
|
|
|
opad[1] = packv (tmps, opad, gid, 1);
|
|
|
|
opad[2] = packv (tmps, opad, gid, 2);
|
|
|
|
opad[3] = packv (tmps, opad, gid, 3);
|
|
|
|
opad[4] = packv (tmps, opad, gid, 4);
|
|
|
|
|
|
|
|
for (u32 i = 0; i < 8; i += 5)
|
|
|
|
{
|
|
|
|
u32x dgst[5];
|
|
|
|
u32x out [5];
|
|
|
|
|
|
|
|
dgst[0] = packv (tmps, dgst, gid, i + 0);
|
|
|
|
dgst[1] = packv (tmps, dgst, gid, i + 1);
|
|
|
|
dgst[2] = packv (tmps, dgst, gid, i + 2);
|
|
|
|
dgst[3] = packv (tmps, dgst, gid, i + 3);
|
|
|
|
dgst[4] = packv (tmps, dgst, gid, i + 4);
|
|
|
|
|
|
|
|
out[0] = packv (tmps, out, gid, i + 0);
|
|
|
|
out[1] = packv (tmps, out, gid, i + 1);
|
|
|
|
out[2] = packv (tmps, out, gid, i + 2);
|
|
|
|
out[3] = packv (tmps, out, gid, i + 3);
|
|
|
|
out[4] = packv (tmps, out, gid, i + 4);
|
|
|
|
|
|
|
|
for (u32 j = 0; j < LOOP_CNT; j++)
|
|
|
|
{
|
|
|
|
u32x w0[4];
|
|
|
|
u32x w1[4];
|
|
|
|
u32x w2[4];
|
|
|
|
u32x w3[4];
|
|
|
|
|
|
|
|
w0[0] = dgst[0];
|
|
|
|
w0[1] = dgst[1];
|
|
|
|
w0[2] = dgst[2];
|
|
|
|
w0[3] = dgst[3];
|
|
|
|
w1[0] = dgst[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;
|
|
|
|
|
|
|
|
hmac_sha1_run_V (w0, w1, w2, w3, ipad, opad, dgst);
|
|
|
|
|
|
|
|
out[0] ^= dgst[0];
|
|
|
|
out[1] ^= dgst[1];
|
|
|
|
out[2] ^= dgst[2];
|
|
|
|
out[3] ^= dgst[3];
|
|
|
|
out[4] ^= dgst[4];
|
|
|
|
}
|
|
|
|
|
|
|
|
unpackv (tmps, dgst, gid, i + 0, dgst[0]);
|
|
|
|
unpackv (tmps, dgst, gid, i + 1, dgst[1]);
|
|
|
|
unpackv (tmps, dgst, gid, i + 2, dgst[2]);
|
|
|
|
unpackv (tmps, dgst, gid, i + 3, dgst[3]);
|
|
|
|
unpackv (tmps, dgst, gid, i + 4, dgst[4]);
|
|
|
|
|
|
|
|
unpackv (tmps, out, gid, i + 0, out[0]);
|
|
|
|
unpackv (tmps, out, gid, i + 1, out[1]);
|
|
|
|
unpackv (tmps, out, gid, i + 2, out[2]);
|
|
|
|
unpackv (tmps, out, gid, i + 3, out[3]);
|
|
|
|
unpackv (tmps, out, gid, i + 4, out[4]);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
KERNEL_FQ void m23100_comp (KERN_ATTR_TMPS_ESALT (keychain_tmp_t, keychain_t))
|
|
|
|
{
|
|
|
|
/**
|
|
|
|
* base
|
|
|
|
*/
|
|
|
|
|
|
|
|
const u64 gid = get_global_id (0);
|
|
|
|
const u64 lid = get_local_id (0);
|
|
|
|
const u64 lsz = get_local_size (0);
|
|
|
|
|
|
|
|
/**
|
|
|
|
* des shared
|
|
|
|
*/
|
|
|
|
|
|
|
|
#ifdef REAL_SHM
|
|
|
|
|
|
|
|
LOCAL_VK u32 s_SPtrans[8][64];
|
|
|
|
LOCAL_VK u32 s_skb[8][64];
|
|
|
|
|
|
|
|
for (u32 i = lid; i < 64; i += lsz)
|
|
|
|
{
|
|
|
|
s_SPtrans[0][i] = c_SPtrans[0][i];
|
|
|
|
s_SPtrans[1][i] = c_SPtrans[1][i];
|
|
|
|
s_SPtrans[2][i] = c_SPtrans[2][i];
|
|
|
|
s_SPtrans[3][i] = c_SPtrans[3][i];
|
|
|
|
s_SPtrans[4][i] = c_SPtrans[4][i];
|
|
|
|
s_SPtrans[5][i] = c_SPtrans[5][i];
|
|
|
|
s_SPtrans[6][i] = c_SPtrans[6][i];
|
|
|
|
s_SPtrans[7][i] = c_SPtrans[7][i];
|
|
|
|
|
|
|
|
s_skb[0][i] = c_skb[0][i];
|
|
|
|
s_skb[1][i] = c_skb[1][i];
|
|
|
|
s_skb[2][i] = c_skb[2][i];
|
|
|
|
s_skb[3][i] = c_skb[3][i];
|
|
|
|
s_skb[4][i] = c_skb[4][i];
|
|
|
|
s_skb[5][i] = c_skb[5][i];
|
|
|
|
s_skb[6][i] = c_skb[6][i];
|
|
|
|
s_skb[7][i] = c_skb[7][i];
|
|
|
|
}
|
|
|
|
|
|
|
|
SYNC_THREADS ();
|
|
|
|
|
|
|
|
#else
|
|
|
|
|
|
|
|
CONSTANT_AS u32a (*s_SPtrans)[64] = c_SPtrans;
|
|
|
|
CONSTANT_AS u32a (*s_skb)[64] = c_skb;
|
|
|
|
|
|
|
|
#endif
|
|
|
|
|
|
|
|
if (gid >= GID_CNT) return;
|
|
|
|
|
|
|
|
u32 key[6];
|
|
|
|
|
|
|
|
key[0] = tmps[gid].out[0];
|
|
|
|
key[1] = tmps[gid].out[1];
|
|
|
|
key[2] = tmps[gid].out[2];
|
|
|
|
key[3] = tmps[gid].out[3];
|
|
|
|
key[4] = tmps[gid].out[4];
|
|
|
|
key[5] = tmps[gid].out[5];
|
|
|
|
|
|
|
|
key[0] = hc_swap32_S (key[0]);
|
|
|
|
key[1] = hc_swap32_S (key[1]);
|
|
|
|
key[2] = hc_swap32_S (key[2]);
|
|
|
|
key[3] = hc_swap32_S (key[3]);
|
|
|
|
key[4] = hc_swap32_S (key[4]);
|
|
|
|
key[5] = hc_swap32_S (key[5]);
|
|
|
|
|
|
|
|
u32 iv[2];
|
|
|
|
|
|
|
|
iv[0] = esalt_bufs[DIGESTS_OFFSET_HOST].data[8];
|
|
|
|
iv[1] = esalt_bufs[DIGESTS_OFFSET_HOST].data[9];
|
|
|
|
|
|
|
|
u32 data[2];
|
|
|
|
|
|
|
|
data[0] = esalt_bufs[DIGESTS_OFFSET_HOST].data[10];
|
|
|
|
data[1] = esalt_bufs[DIGESTS_OFFSET_HOST].data[11];
|
|
|
|
|
|
|
|
// 3DES-CBC (decrypt, encrypt, decrypt):
|
|
|
|
|
|
|
|
u32 Ka[16];
|
|
|
|
u32 Kb[16];
|
|
|
|
u32 out[2];
|
|
|
|
|
|
|
|
// 1 (decrypt)
|
|
|
|
|
|
|
|
_des_crypt_keysetup (key[4], key[5], Ka, Kb, s_skb);
|
|
|
|
_des_crypt_decrypt (out, data, Ka, Kb, s_SPtrans);
|
|
|
|
|
|
|
|
// 2 (encrypt)
|
|
|
|
|
|
|
|
_des_crypt_keysetup (key[2], key[3], Ka, Kb, s_skb);
|
|
|
|
_des_crypt_encrypt (out, out, Ka, Kb, s_SPtrans);
|
|
|
|
|
|
|
|
// 3 (decrypt)
|
|
|
|
|
|
|
|
_des_crypt_keysetup (key[0], key[1], Ka, Kb, s_skb);
|
|
|
|
_des_crypt_decrypt (out, out, Ka, Kb, s_SPtrans);
|
|
|
|
|
|
|
|
if ((out[1] ^ iv[1]) == 0x04040404) // this check uses very low number of bits => collisions
|
|
|
|
{
|
|
|
|
if (hc_atomic_inc (&hashes_shown[DIGESTS_OFFSET_HOST]) == 0)
|
|
|
|
{
|
|
|
|
mark_hash (plains_buf, d_return_buf, SALT_POS_HOST, DIGESTS_CNT, 0, DIGESTS_OFFSET_HOST + 0, gid, 0, 0, 0);
|
|
|
|
}
|
|
|
|
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
}
|