mirror of
https://github.com/hashcat/hashcat.git
synced 2024-12-22 14:48:12 +00:00
Converted to new SIMD: -m 5500 -a 0
This commit is contained in:
Jens Steube
parent
b3a43b47a1
commit
ca7d382dce
@ -7,6 +7,8 @@
|
||||
|
||||
#define _MD4_
|
||||
|
||||
#define NEW_SIMD_CODE
|
||||
|
||||
#include "include/constants.h"
|
||||
#include "include/kernel_vendor.h"
|
||||
|
||||
@ -20,9 +22,7 @@
|
||||
#include "OpenCL/common.c"
|
||||
#include "include/rp_kernel.h"
|
||||
#include "OpenCL/rp.c"
|
||||
|
||||
#define COMPARE_S "OpenCL/check_single_comp4.c"
|
||||
#define COMPARE_M "OpenCL/check_multi_comp4.c"
|
||||
#include "OpenCL/simd.c"
|
||||
|
||||
#define PERM_OP(a,b,tt,n,m) \
|
||||
{ \
|
||||
@ -340,18 +340,26 @@ __constant u32 c_skb[8][64] =
|
||||
}
|
||||
};
|
||||
|
||||
#if VECT_SIZE == 1
|
||||
#define BOX(i,n,S) (S)[(n)][(i)]
|
||||
#elif VECT_SIZE == 2
|
||||
#define BOX(i,n,S) (u32x) ((S)[(n)][(i).s0], (S)[(n)][(i).s1])
|
||||
#elif VECT_SIZE == 4
|
||||
#define BOX(i,n,S) (u32x) ((S)[(n)][(i).s0], (S)[(n)][(i).s1], (S)[(n)][(i).s2], (S)[(n)][(i).s3])
|
||||
#elif VECT_SIZE == 8
|
||||
#define BOX(i,n,S) (u32x) ((S)[(n)][(i).s0], (S)[(n)][(i).s1], (S)[(n)][(i).s2], (S)[(n)][(i).s3], (S)[(n)][(i).s4], (S)[(n)][(i).s5], (S)[(n)][(i).s6], (S)[(n)][(i).s7])
|
||||
#endif
|
||||
|
||||
static void _des_crypt_encrypt (u32 iv[2], u32 data[2], u32 Kc[16], u32 Kd[16], __local u32 (*s_SPtrans)[64])
|
||||
static void _des_crypt_encrypt (u32x iv[2], u32x data[2], u32x Kc[16], u32x Kd[16], __local u32 (*s_SPtrans)[64])
|
||||
{
|
||||
u32 r = data[0];
|
||||
u32 l = data[1];
|
||||
u32x r = data[0];
|
||||
u32x l = data[1];
|
||||
|
||||
#pragma unroll 16
|
||||
for (u32 i = 0; i < 16; i += 2)
|
||||
{
|
||||
u32 u;
|
||||
u32 t;
|
||||
u32x u;
|
||||
u32x t;
|
||||
|
||||
u = Kc[i + 0] ^ rotl32 (r, 30u);
|
||||
t = Kd[i + 0] ^ rotl32 (r, 26u);
|
||||
@ -382,9 +390,9 @@ static void _des_crypt_encrypt (u32 iv[2], u32 data[2], u32 Kc[16], u32 Kd[16],
|
||||
iv[1] = r;
|
||||
}
|
||||
|
||||
static void _des_crypt_keysetup (u32 c, u32 d, u32 Kc[16], u32 Kd[16], __local u32 (*s_skb)[64])
|
||||
static void _des_crypt_keysetup (u32x c, u32x d, u32x Kc[16], u32x Kd[16], __local u32 (*s_skb)[64])
|
||||
{
|
||||
u32 tt;
|
||||
u32x tt;
|
||||
|
||||
PERM_OP (d, c, tt, 4, 0x0f0f0f0f);
|
||||
HPERM_OP (c, tt, 2, 0xcccc0000);
|
||||
@ -417,32 +425,32 @@ static void _des_crypt_keysetup (u32 c, u32 d, u32 Kc[16], u32 Kd[16], __local u
|
||||
c = c & 0x0fffffff;
|
||||
d = d & 0x0fffffff;
|
||||
|
||||
const u32 c00 = (c >> 0) & 0x0000003f;
|
||||
const u32 c06 = (c >> 6) & 0x00383003;
|
||||
const u32 c07 = (c >> 7) & 0x0000003c;
|
||||
const u32 c13 = (c >> 13) & 0x0000060f;
|
||||
const u32 c20 = (c >> 20) & 0x00000001;
|
||||
const u32x c00 = (c >> 0) & 0x0000003f;
|
||||
const u32x c06 = (c >> 6) & 0x00383003;
|
||||
const u32x c07 = (c >> 7) & 0x0000003c;
|
||||
const u32x c13 = (c >> 13) & 0x0000060f;
|
||||
const u32x c20 = (c >> 20) & 0x00000001;
|
||||
|
||||
u32 s = BOX (((c00 >> 0) & 0xff), 0, s_skb)
|
||||
| BOX (((c06 >> 0) & 0xff)
|
||||
|((c07 >> 0) & 0xff), 1, s_skb)
|
||||
| BOX (((c13 >> 0) & 0xff)
|
||||
|((c06 >> 8) & 0xff), 2, s_skb)
|
||||
| BOX (((c20 >> 0) & 0xff)
|
||||
|((c13 >> 8) & 0xff)
|
||||
|((c06 >> 16) & 0xff), 3, s_skb);
|
||||
u32x s = BOX (((c00 >> 0) & 0xff), 0, s_skb)
|
||||
| BOX (((c06 >> 0) & 0xff)
|
||||
|((c07 >> 0) & 0xff), 1, s_skb)
|
||||
| BOX (((c13 >> 0) & 0xff)
|
||||
|((c06 >> 8) & 0xff), 2, s_skb)
|
||||
| BOX (((c20 >> 0) & 0xff)
|
||||
|((c13 >> 8) & 0xff)
|
||||
|((c06 >> 16) & 0xff), 3, s_skb);
|
||||
|
||||
const u32 d00 = (d >> 0) & 0x00003c3f;
|
||||
const u32 d07 = (d >> 7) & 0x00003f03;
|
||||
const u32 d21 = (d >> 21) & 0x0000000f;
|
||||
const u32 d22 = (d >> 22) & 0x00000030;
|
||||
const u32x d00 = (d >> 0) & 0x00003c3f;
|
||||
const u32x d07 = (d >> 7) & 0x00003f03;
|
||||
const u32x d21 = (d >> 21) & 0x0000000f;
|
||||
const u32x d22 = (d >> 22) & 0x00000030;
|
||||
|
||||
u32 t = BOX (((d00 >> 0) & 0xff), 4, s_skb)
|
||||
| BOX (((d07 >> 0) & 0xff)
|
||||
|((d00 >> 8) & 0xff), 5, s_skb)
|
||||
| BOX (((d07 >> 8) & 0xff), 6, s_skb)
|
||||
| BOX (((d21 >> 0) & 0xff)
|
||||
|((d22 >> 0) & 0xff), 7, s_skb);
|
||||
u32x t = BOX (((d00 >> 0) & 0xff), 4, s_skb)
|
||||
| BOX (((d07 >> 0) & 0xff)
|
||||
|((d00 >> 8) & 0xff), 5, s_skb)
|
||||
| BOX (((d07 >> 8) & 0xff), 6, s_skb)
|
||||
| BOX (((d21 >> 0) & 0xff)
|
||||
|((d22 >> 0) & 0xff), 7, s_skb);
|
||||
|
||||
Kc[i] = ((t << 16) | (s & 0x0000ffff));
|
||||
Kd[i] = ((s >> 16) | (t & 0xffff0000));
|
||||
@ -563,54 +571,31 @@ __kernel void m05500_m04 (__global pw_t *pws, __global kernel_rule_t * rules_bu
|
||||
* loop
|
||||
*/
|
||||
|
||||
for (u32 il_pos = 0; il_pos < rules_cnt; il_pos++)
|
||||
for (u32 il_pos = 0; il_pos < rules_cnt; il_pos += VECT_SIZE)
|
||||
{
|
||||
u32 w0[4];
|
||||
u32x w0[4] = { 0 };
|
||||
u32x w1[4] = { 0 };
|
||||
u32x w2[4] = { 0 };
|
||||
u32x w3[4] = { 0 };
|
||||
|
||||
w0[0] = pw_buf0[0];
|
||||
w0[1] = pw_buf0[1];
|
||||
w0[2] = pw_buf0[2];
|
||||
w0[3] = pw_buf0[3];
|
||||
|
||||
u32 w1[4];
|
||||
|
||||
w1[0] = pw_buf1[0];
|
||||
w1[1] = pw_buf1[1];
|
||||
w1[2] = pw_buf1[2];
|
||||
w1[3] = pw_buf1[3];
|
||||
|
||||
u32 w2[4];
|
||||
|
||||
w2[0] = 0;
|
||||
w2[1] = 0;
|
||||
w2[2] = 0;
|
||||
w2[3] = 0;
|
||||
|
||||
u32 w3[4];
|
||||
|
||||
w3[0] = 0;
|
||||
w3[1] = 0;
|
||||
w3[2] = 0;
|
||||
w3[3] = 0;
|
||||
|
||||
const u32 out_len = apply_rules (rules_buf[il_pos].cmds, w0, w1, pw_len);
|
||||
const u32 out_len = apply_rules_vect (pw_buf0, pw_buf1, pw_len, rules_buf, il_pos, w0, w1);
|
||||
|
||||
append_0x80_2x4 (w0, w1, out_len);
|
||||
|
||||
u32 w0_t[4];
|
||||
u32 w1_t[4];
|
||||
u32 w2_t[4];
|
||||
u32 w3_t[4];
|
||||
u32x w0_t[4];
|
||||
u32x w1_t[4];
|
||||
u32x w2_t[4];
|
||||
u32x w3_t[4];
|
||||
|
||||
make_unicode (w0, w0_t, w1_t);
|
||||
make_unicode (w1, w2_t, w3_t);
|
||||
|
||||
w3_t[2] = out_len * 8 * 2;
|
||||
|
||||
u32 a = MD4M_A;
|
||||
u32 b = MD4M_B;
|
||||
u32 c = MD4M_C;
|
||||
u32 d = MD4M_D;
|
||||
u32x a = MD4M_A;
|
||||
u32x b = MD4M_B;
|
||||
u32x c = MD4M_C;
|
||||
u32x d = MD4M_D;
|
||||
|
||||
MD4_STEP (MD4_Fo, a, b, c, d, w0_t[0], MD4C00, MD4S00);
|
||||
MD4_STEP (MD4_Fo, d, a, b, c, w0_t[1], MD4C00, MD4S01);
|
||||
@ -661,7 +646,7 @@ __kernel void m05500_m04 (__global pw_t *pws, __global kernel_rule_t * rules_bu
|
||||
MD4_STEP (MD4_H , a, b, c, d, w0_t[3], MD4C02, MD4S20);
|
||||
MD4_STEP (MD4_H , d, a, b, c, w2_t[3], MD4C02, MD4S21);
|
||||
|
||||
if (allx (s2 != ((d + MD4M_D) >> 16))) continue;
|
||||
if (MATCHES_NONE_VS (((d + MD4M_D) >> 16), s2)) continue;
|
||||
|
||||
MD4_STEP (MD4_H , c, d, a, b, w1_t[3], MD4C02, MD4S22);
|
||||
MD4_STEP (MD4_H , b, c, d, a, w3_t[3], MD4C02, MD4S23);
|
||||
@ -675,16 +660,16 @@ __kernel void m05500_m04 (__global pw_t *pws, __global kernel_rule_t * rules_bu
|
||||
* DES1
|
||||
*/
|
||||
|
||||
u32 key[2];
|
||||
u32x key[2];
|
||||
|
||||
transform_netntlmv1_key (a, b, key);
|
||||
|
||||
u32 Kc[16];
|
||||
u32 Kd[16];
|
||||
u32x Kc[16];
|
||||
u32x Kd[16];
|
||||
|
||||
_des_crypt_keysetup (key[0], key[1], Kc, Kd, s_skb);
|
||||
|
||||
volatile u32 iv1[2];
|
||||
u32x iv1[2];
|
||||
|
||||
_des_crypt_encrypt (iv1, data, Kc, Kd, s_SPtrans);
|
||||
|
||||
@ -692,14 +677,14 @@ __kernel void m05500_m04 (__global pw_t *pws, __global kernel_rule_t * rules_bu
|
||||
* DES2
|
||||
*/
|
||||
|
||||
const u32 bc = (b >> 24) | (c << 8);
|
||||
const u32 cd = (c >> 24) | (d << 8);
|
||||
const u32x bc = (b >> 24) | (c << 8);
|
||||
const u32x cd = (c >> 24) | (d << 8);
|
||||
|
||||
transform_netntlmv1_key (bc, cd, key);
|
||||
|
||||
_des_crypt_keysetup (key[0], key[1], Kc, Kd, s_skb);
|
||||
|
||||
u32 iv2[2];
|
||||
u32x iv2[2];
|
||||
|
||||
_des_crypt_encrypt (iv2, data, Kc, Kd, s_SPtrans);
|
||||
|
||||
@ -707,12 +692,7 @@ __kernel void m05500_m04 (__global pw_t *pws, __global kernel_rule_t * rules_bu
|
||||
* compare
|
||||
*/
|
||||
|
||||
const u32 r0 = iv1[0];
|
||||
const u32 r1 = iv1[1];
|
||||
const u32 r2 = iv2[0];
|
||||
const u32 r3 = iv2[1];
|
||||
|
||||
#include COMPARE_M
|
||||
COMPARE_M_SIMD (iv1[0], iv1[1], iv2[0], iv2[1]);
|
||||
}
|
||||
}
|
||||
|
||||
@ -815,54 +795,31 @@ __kernel void m05500_s04 (__global pw_t *pws, __global kernel_rule_t * rules_bu
|
||||
* loop
|
||||
*/
|
||||
|
||||
for (u32 il_pos = 0; il_pos < rules_cnt; il_pos++)
|
||||
for (u32 il_pos = 0; il_pos < rules_cnt; il_pos += VECT_SIZE)
|
||||
{
|
||||
u32 w0[4];
|
||||
u32x w0[4] = { 0 };
|
||||
u32x w1[4] = { 0 };
|
||||
u32x w2[4] = { 0 };
|
||||
u32x w3[4] = { 0 };
|
||||
|
||||
w0[0] = pw_buf0[0];
|
||||
w0[1] = pw_buf0[1];
|
||||
w0[2] = pw_buf0[2];
|
||||
w0[3] = pw_buf0[3];
|
||||
|
||||
u32 w1[4];
|
||||
|
||||
w1[0] = pw_buf1[0];
|
||||
w1[1] = pw_buf1[1];
|
||||
w1[2] = pw_buf1[2];
|
||||
w1[3] = pw_buf1[3];
|
||||
|
||||
u32 w2[4];
|
||||
|
||||
w2[0] = 0;
|
||||
w2[1] = 0;
|
||||
w2[2] = 0;
|
||||
w2[3] = 0;
|
||||
|
||||
u32 w3[4];
|
||||
|
||||
w3[0] = 0;
|
||||
w3[1] = 0;
|
||||
w3[2] = 0;
|
||||
w3[3] = 0;
|
||||
|
||||
const u32 out_len = apply_rules (rules_buf[il_pos].cmds, w0, w1, pw_len);
|
||||
const u32 out_len = apply_rules_vect (pw_buf0, pw_buf1, pw_len, rules_buf, il_pos, w0, w1);
|
||||
|
||||
append_0x80_2x4 (w0, w1, out_len);
|
||||
|
||||
u32 w0_t[4];
|
||||
u32 w1_t[4];
|
||||
u32 w2_t[4];
|
||||
u32 w3_t[4];
|
||||
u32x w0_t[4];
|
||||
u32x w1_t[4];
|
||||
u32x w2_t[4];
|
||||
u32x w3_t[4];
|
||||
|
||||
make_unicode (w0, w0_t, w1_t);
|
||||
make_unicode (w1, w2_t, w3_t);
|
||||
|
||||
w3_t[2] = out_len * 8 * 2;
|
||||
|
||||
u32 a = MD4M_A;
|
||||
u32 b = MD4M_B;
|
||||
u32 c = MD4M_C;
|
||||
u32 d = MD4M_D;
|
||||
u32x a = MD4M_A;
|
||||
u32x b = MD4M_B;
|
||||
u32x c = MD4M_C;
|
||||
u32x d = MD4M_D;
|
||||
|
||||
MD4_STEP (MD4_Fo, a, b, c, d, w0_t[0], MD4C00, MD4S00);
|
||||
MD4_STEP (MD4_Fo, d, a, b, c, w0_t[1], MD4C00, MD4S01);
|
||||
@ -913,7 +870,7 @@ __kernel void m05500_s04 (__global pw_t *pws, __global kernel_rule_t * rules_bu
|
||||
MD4_STEP (MD4_H , a, b, c, d, w0_t[3], MD4C02, MD4S20);
|
||||
MD4_STEP (MD4_H , d, a, b, c, w2_t[3], MD4C02, MD4S21);
|
||||
|
||||
if (allx (s2 != ((d + MD4M_D) >> 16))) continue;
|
||||
if (MATCHES_NONE_VS (((d + MD4M_D) >> 16), s2)) continue;
|
||||
|
||||
MD4_STEP (MD4_H , c, d, a, b, w1_t[3], MD4C02, MD4S22);
|
||||
MD4_STEP (MD4_H , b, c, d, a, w3_t[3], MD4C02, MD4S23);
|
||||
@ -927,16 +884,16 @@ __kernel void m05500_s04 (__global pw_t *pws, __global kernel_rule_t * rules_bu
|
||||
* DES1
|
||||
*/
|
||||
|
||||
u32 key[2];
|
||||
u32x key[2];
|
||||
|
||||
transform_netntlmv1_key (a, b, key);
|
||||
|
||||
u32 Kc[16];
|
||||
u32 Kd[16];
|
||||
u32x Kc[16];
|
||||
u32x Kd[16];
|
||||
|
||||
_des_crypt_keysetup (key[0], key[1], Kc, Kd, s_skb);
|
||||
|
||||
u32 iv1[2];
|
||||
u32x iv1[2];
|
||||
|
||||
_des_crypt_encrypt (iv1, data, Kc, Kd, s_SPtrans);
|
||||
|
||||
@ -954,7 +911,7 @@ __kernel void m05500_s04 (__global pw_t *pws, __global kernel_rule_t * rules_bu
|
||||
_des_crypt_encrypt (iv2, data, Kc, Kd, s_SPtrans);
|
||||
*/
|
||||
|
||||
u32 iv2[2];
|
||||
u32x iv2[2];
|
||||
|
||||
iv2[0] = search[2];
|
||||
iv2[1] = search[3];
|
||||
@ -963,12 +920,7 @@ __kernel void m05500_s04 (__global pw_t *pws, __global kernel_rule_t * rules_bu
|
||||
* compare
|
||||
*/
|
||||
|
||||
const u32 r0 = iv1[0];
|
||||
const u32 r1 = iv1[1];
|
||||
const u32 r2 = iv2[0];
|
||||
const u32 r3 = iv2[1];
|
||||
|
||||
#include COMPARE_S
|
||||
COMPARE_S_SIMD (iv1[0], iv1[1], iv2[0], iv2[1]);
|
||||
}
|
||||
}
|
||||
|
||||
|
||||
Loading…
Reference in New Issue
Block a user