mirror of
https://github.com/hashcat/hashcat.git
synced 2024-11-22 08:08:10 +00:00
589 lines
19 KiB
Common Lisp
589 lines
19 KiB
Common Lisp
/**
|
|
* Author......: See docs/credits.txt
|
|
* License.....: MIT
|
|
*/
|
|
|
|
#define NEW_SIMD_CODE
|
|
|
|
#ifdef KERNEL_STATIC
|
|
#include M2S(INCLUDE_PATH/inc_vendor.h)
|
|
#include M2S(INCLUDE_PATH/inc_types.h)
|
|
#include M2S(INCLUDE_PATH/inc_platform.cl)
|
|
#include M2S(INCLUDE_PATH/inc_common.cl)
|
|
#include M2S(INCLUDE_PATH/inc_simd.cl)
|
|
#include M2S(INCLUDE_PATH/inc_hash_md4.cl)
|
|
#endif
|
|
|
|
#define MD4_STEP_REV(f,a,b,c,d,x,t,s) \
|
|
{ \
|
|
a = hc_rotr32_S (a, s); \
|
|
a -= f (b, c, d); \
|
|
a -= x; \
|
|
a -= t; \
|
|
}
|
|
|
|
#define MD4_STEP_REV1(f,a,b,c,d,x,t,s) \
|
|
{ \
|
|
a = hc_rotr32_S (a, s); \
|
|
a -= x; \
|
|
a -= t; \
|
|
}
|
|
|
|
DECLSPEC void m01000m (PRIVATE_AS u32 *w, const u32 pw_len, KERN_ATTR_FUNC_VECTOR ())
|
|
{
|
|
/**
|
|
* modifiers are taken from args
|
|
*/
|
|
|
|
/**
|
|
* base
|
|
*/
|
|
|
|
const u32 F_w0c00 = 0 + MD4C00;
|
|
const u32 F_w1c00 = w[ 1] + MD4C00;
|
|
const u32 F_w2c00 = w[ 2] + MD4C00;
|
|
const u32 F_w3c00 = w[ 3] + MD4C00;
|
|
const u32 F_w4c00 = w[ 4] + MD4C00;
|
|
const u32 F_w5c00 = w[ 5] + MD4C00;
|
|
const u32 F_w6c00 = w[ 6] + MD4C00;
|
|
const u32 F_w7c00 = w[ 7] + MD4C00;
|
|
const u32 F_w8c00 = w[ 8] + MD4C00;
|
|
const u32 F_w9c00 = w[ 9] + MD4C00;
|
|
const u32 F_wac00 = w[10] + MD4C00;
|
|
const u32 F_wbc00 = w[11] + MD4C00;
|
|
const u32 F_wcc00 = w[12] + MD4C00;
|
|
const u32 F_wdc00 = w[13] + MD4C00;
|
|
const u32 F_wec00 = w[14] + MD4C00;
|
|
const u32 F_wfc00 = w[15] + MD4C00;
|
|
|
|
const u32 G_w0c01 = 0 + MD4C01;
|
|
const u32 G_w4c01 = w[ 4] + MD4C01;
|
|
const u32 G_w8c01 = w[ 8] + MD4C01;
|
|
const u32 G_wcc01 = w[12] + MD4C01;
|
|
const u32 G_w1c01 = w[ 1] + MD4C01;
|
|
const u32 G_w5c01 = w[ 5] + MD4C01;
|
|
const u32 G_w9c01 = w[ 9] + MD4C01;
|
|
const u32 G_wdc01 = w[13] + MD4C01;
|
|
const u32 G_w2c01 = w[ 2] + MD4C01;
|
|
const u32 G_w6c01 = w[ 6] + MD4C01;
|
|
const u32 G_wac01 = w[10] + MD4C01;
|
|
const u32 G_wec01 = w[14] + MD4C01;
|
|
const u32 G_w3c01 = w[ 3] + MD4C01;
|
|
const u32 G_w7c01 = w[ 7] + MD4C01;
|
|
const u32 G_wbc01 = w[11] + MD4C01;
|
|
const u32 G_wfc01 = w[15] + MD4C01;
|
|
|
|
const u32 H_w0c02 = 0 + MD4C02;
|
|
const u32 H_w8c02 = w[ 8] + MD4C02;
|
|
const u32 H_w4c02 = w[ 4] + MD4C02;
|
|
const u32 H_wcc02 = w[12] + MD4C02;
|
|
const u32 H_w2c02 = w[ 2] + MD4C02;
|
|
const u32 H_wac02 = w[10] + MD4C02;
|
|
const u32 H_w6c02 = w[ 6] + MD4C02;
|
|
const u32 H_wec02 = w[14] + MD4C02;
|
|
const u32 H_w1c02 = w[ 1] + MD4C02;
|
|
const u32 H_w9c02 = w[ 9] + MD4C02;
|
|
const u32 H_w5c02 = w[ 5] + MD4C02;
|
|
const u32 H_wdc02 = w[13] + MD4C02;
|
|
const u32 H_w3c02 = w[ 3] + MD4C02;
|
|
const u32 H_wbc02 = w[11] + MD4C02;
|
|
const u32 H_w7c02 = w[ 7] + MD4C02;
|
|
const u32 H_wfc02 = w[15] + MD4C02;
|
|
|
|
/**
|
|
* loop
|
|
*/
|
|
|
|
u32 w0l = w[0];
|
|
|
|
for (u32 il_pos = 0; il_pos < IL_CNT; il_pos += VECT_SIZE)
|
|
{
|
|
const u32x w0r = words_buf_r[il_pos / VECT_SIZE];
|
|
|
|
const u32x w0 = w0l | w0r;
|
|
|
|
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, F_w0c00, MD4S00);
|
|
MD4_STEP0(MD4_Fo, d, a, b, c, F_w1c00, MD4S01);
|
|
MD4_STEP0(MD4_Fo, c, d, a, b, F_w2c00, MD4S02);
|
|
MD4_STEP0(MD4_Fo, b, c, d, a, F_w3c00, MD4S03);
|
|
MD4_STEP0(MD4_Fo, a, b, c, d, F_w4c00, MD4S00);
|
|
MD4_STEP0(MD4_Fo, d, a, b, c, F_w5c00, MD4S01);
|
|
MD4_STEP0(MD4_Fo, c, d, a, b, F_w6c00, MD4S02);
|
|
MD4_STEP0(MD4_Fo, b, c, d, a, F_w7c00, MD4S03);
|
|
MD4_STEP0(MD4_Fo, a, b, c, d, F_w8c00, MD4S00);
|
|
MD4_STEP0(MD4_Fo, d, a, b, c, F_w9c00, MD4S01);
|
|
MD4_STEP0(MD4_Fo, c, d, a, b, F_wac00, MD4S02);
|
|
MD4_STEP0(MD4_Fo, b, c, d, a, F_wbc00, MD4S03);
|
|
MD4_STEP0(MD4_Fo, a, b, c, d, F_wcc00, MD4S00);
|
|
MD4_STEP0(MD4_Fo, d, a, b, c, F_wdc00, MD4S01);
|
|
MD4_STEP0(MD4_Fo, c, d, a, b, F_wec00, MD4S02);
|
|
MD4_STEP0(MD4_Fo, b, c, d, a, F_wfc00, MD4S03);
|
|
|
|
MD4_STEP (MD4_Go, a, b, c, d, w0, G_w0c01, MD4S10);
|
|
MD4_STEP0(MD4_Go, d, a, b, c, G_w4c01, MD4S11);
|
|
MD4_STEP0(MD4_Go, c, d, a, b, G_w8c01, MD4S12);
|
|
MD4_STEP0(MD4_Go, b, c, d, a, G_wcc01, MD4S13);
|
|
MD4_STEP0(MD4_Go, a, b, c, d, G_w1c01, MD4S10);
|
|
MD4_STEP0(MD4_Go, d, a, b, c, G_w5c01, MD4S11);
|
|
MD4_STEP0(MD4_Go, c, d, a, b, G_w9c01, MD4S12);
|
|
MD4_STEP0(MD4_Go, b, c, d, a, G_wdc01, MD4S13);
|
|
MD4_STEP0(MD4_Go, a, b, c, d, G_w2c01, MD4S10);
|
|
MD4_STEP0(MD4_Go, d, a, b, c, G_w6c01, MD4S11);
|
|
MD4_STEP0(MD4_Go, c, d, a, b, G_wac01, MD4S12);
|
|
MD4_STEP0(MD4_Go, b, c, d, a, G_wec01, MD4S13);
|
|
MD4_STEP0(MD4_Go, a, b, c, d, G_w3c01, MD4S10);
|
|
MD4_STEP0(MD4_Go, d, a, b, c, G_w7c01, MD4S11);
|
|
MD4_STEP0(MD4_Go, c, d, a, b, G_wbc01, MD4S12);
|
|
MD4_STEP0(MD4_Go, b, c, d, a, G_wfc01, MD4S13);
|
|
|
|
MD4_STEP (MD4_H , a, b, c, d, w0, H_w0c02, MD4S20);
|
|
MD4_STEP0(MD4_H , d, a, b, c, H_w8c02, MD4S21);
|
|
MD4_STEP0(MD4_H , c, d, a, b, H_w4c02, MD4S22);
|
|
MD4_STEP0(MD4_H , b, c, d, a, H_wcc02, MD4S23);
|
|
MD4_STEP0(MD4_H , a, b, c, d, H_w2c02, MD4S20);
|
|
MD4_STEP0(MD4_H , d, a, b, c, H_wac02, MD4S21);
|
|
MD4_STEP0(MD4_H , c, d, a, b, H_w6c02, MD4S22);
|
|
MD4_STEP0(MD4_H , b, c, d, a, H_wec02, MD4S23);
|
|
MD4_STEP0(MD4_H , a, b, c, d, H_w1c02, MD4S20);
|
|
MD4_STEP0(MD4_H , d, a, b, c, H_w9c02, MD4S21);
|
|
MD4_STEP0(MD4_H , c, d, a, b, H_w5c02, MD4S22);
|
|
MD4_STEP0(MD4_H , b, c, d, a, H_wdc02, MD4S23);
|
|
MD4_STEP0(MD4_H , a, b, c, d, H_w3c02, MD4S20);
|
|
MD4_STEP0(MD4_H , d, a, b, c, H_wbc02, MD4S21);
|
|
MD4_STEP0(MD4_H , c, d, a, b, H_w7c02, MD4S22);
|
|
MD4_STEP0(MD4_H , b, c, d, a, H_wfc02, MD4S23);
|
|
|
|
COMPARE_M_SIMD (a, d, c, b);
|
|
}
|
|
}
|
|
|
|
DECLSPEC void m01000s (PRIVATE_AS u32 *w, const u32 pw_len, KERN_ATTR_FUNC_VECTOR ())
|
|
{
|
|
/**
|
|
* modifiers are taken from args
|
|
*/
|
|
|
|
/**
|
|
* base
|
|
*/
|
|
|
|
const u32 F_w0c00 = 0 + MD4C00;
|
|
const u32 F_w1c00 = w[ 1] + MD4C00;
|
|
const u32 F_w2c00 = w[ 2] + MD4C00;
|
|
const u32 F_w3c00 = w[ 3] + MD4C00;
|
|
const u32 F_w4c00 = w[ 4] + MD4C00;
|
|
const u32 F_w5c00 = w[ 5] + MD4C00;
|
|
const u32 F_w6c00 = w[ 6] + MD4C00;
|
|
const u32 F_w7c00 = w[ 7] + MD4C00;
|
|
const u32 F_w8c00 = w[ 8] + MD4C00;
|
|
const u32 F_w9c00 = w[ 9] + MD4C00;
|
|
const u32 F_wac00 = w[10] + MD4C00;
|
|
const u32 F_wbc00 = w[11] + MD4C00;
|
|
const u32 F_wcc00 = w[12] + MD4C00;
|
|
const u32 F_wdc00 = w[13] + MD4C00;
|
|
const u32 F_wec00 = w[14] + MD4C00;
|
|
const u32 F_wfc00 = w[15] + MD4C00;
|
|
|
|
const u32 G_w0c01 = 0 + MD4C01;
|
|
const u32 G_w4c01 = w[ 4] + MD4C01;
|
|
const u32 G_w8c01 = w[ 8] + MD4C01;
|
|
const u32 G_wcc01 = w[12] + MD4C01;
|
|
const u32 G_w1c01 = w[ 1] + MD4C01;
|
|
const u32 G_w5c01 = w[ 5] + MD4C01;
|
|
const u32 G_w9c01 = w[ 9] + MD4C01;
|
|
const u32 G_wdc01 = w[13] + MD4C01;
|
|
const u32 G_w2c01 = w[ 2] + MD4C01;
|
|
const u32 G_w6c01 = w[ 6] + MD4C01;
|
|
const u32 G_wac01 = w[10] + MD4C01;
|
|
const u32 G_wec01 = w[14] + MD4C01;
|
|
const u32 G_w3c01 = w[ 3] + MD4C01;
|
|
const u32 G_w7c01 = w[ 7] + MD4C01;
|
|
const u32 G_wbc01 = w[11] + MD4C01;
|
|
const u32 G_wfc01 = w[15] + MD4C01;
|
|
|
|
const u32 H_w0c02 = 0 + MD4C02;
|
|
const u32 H_w8c02 = w[ 8] + MD4C02;
|
|
const u32 H_w4c02 = w[ 4] + MD4C02;
|
|
const u32 H_wcc02 = w[12] + MD4C02;
|
|
const u32 H_w2c02 = w[ 2] + MD4C02;
|
|
const u32 H_wac02 = w[10] + MD4C02;
|
|
const u32 H_w6c02 = w[ 6] + MD4C02;
|
|
const u32 H_wec02 = w[14] + MD4C02;
|
|
const u32 H_w1c02 = w[ 1] + MD4C02;
|
|
const u32 H_w9c02 = w[ 9] + MD4C02;
|
|
const u32 H_w5c02 = w[ 5] + MD4C02;
|
|
const u32 H_wdc02 = w[13] + MD4C02;
|
|
const u32 H_w3c02 = w[ 3] + MD4C02;
|
|
const u32 H_wbc02 = w[11] + MD4C02;
|
|
const u32 H_w7c02 = w[ 7] + MD4C02;
|
|
const u32 H_wfc02 = w[15] + MD4C02;
|
|
|
|
/**
|
|
* digest
|
|
*/
|
|
|
|
const u32 search[4] =
|
|
{
|
|
digests_buf[DIGESTS_OFFSET_HOST].digest_buf[DGST_R0],
|
|
digests_buf[DIGESTS_OFFSET_HOST].digest_buf[DGST_R1],
|
|
digests_buf[DIGESTS_OFFSET_HOST].digest_buf[DGST_R2],
|
|
digests_buf[DIGESTS_OFFSET_HOST].digest_buf[DGST_R3]
|
|
};
|
|
|
|
/**
|
|
* reverse
|
|
*/
|
|
|
|
u32 a_rev = digests_buf[DIGESTS_OFFSET_HOST].digest_buf[0];
|
|
u32 b_rev = digests_buf[DIGESTS_OFFSET_HOST].digest_buf[1];
|
|
u32 c_rev = digests_buf[DIGESTS_OFFSET_HOST].digest_buf[2];
|
|
u32 d_rev = digests_buf[DIGESTS_OFFSET_HOST].digest_buf[3];
|
|
|
|
MD4_STEP_REV (MD4_H_S, b_rev, c_rev, d_rev, a_rev, w[15], MD4C02, MD4S23);
|
|
MD4_STEP_REV (MD4_H_S, c_rev, d_rev, a_rev, b_rev, w[ 7], MD4C02, MD4S22);
|
|
MD4_STEP_REV (MD4_H_S, d_rev, a_rev, b_rev, c_rev, w[11], MD4C02, MD4S21);
|
|
MD4_STEP_REV (MD4_H_S, a_rev, b_rev, c_rev, d_rev, w[ 3], MD4C02, MD4S20);
|
|
MD4_STEP_REV (MD4_H_S, b_rev, c_rev, d_rev, a_rev, w[13], MD4C02, MD4S23);
|
|
MD4_STEP_REV (MD4_H_S, c_rev, d_rev, a_rev, b_rev, w[ 5], MD4C02, MD4S22);
|
|
MD4_STEP_REV (MD4_H_S, d_rev, a_rev, b_rev, c_rev, w[ 9], MD4C02, MD4S21);
|
|
MD4_STEP_REV (MD4_H_S, a_rev, b_rev, c_rev, d_rev, w[ 1], MD4C02, MD4S20);
|
|
MD4_STEP_REV (MD4_H_S, b_rev, c_rev, d_rev, a_rev, w[14], MD4C02, MD4S23);
|
|
MD4_STEP_REV (MD4_H_S, c_rev, d_rev, a_rev, b_rev, w[ 6], MD4C02, MD4S22);
|
|
MD4_STEP_REV (MD4_H_S, d_rev, a_rev, b_rev, c_rev, w[10], MD4C02, MD4S21);
|
|
MD4_STEP_REV (MD4_H_S, a_rev, b_rev, c_rev, d_rev, w[ 2], MD4C02, MD4S20);
|
|
MD4_STEP_REV (MD4_H_S, b_rev, c_rev, d_rev, a_rev, w[12], MD4C02, MD4S23);
|
|
MD4_STEP_REV (MD4_H_S, c_rev, d_rev, a_rev, b_rev, w[ 4], MD4C02, MD4S22);
|
|
MD4_STEP_REV (MD4_H_S, d_rev, a_rev, b_rev, c_rev, w[ 8], MD4C02, MD4S21);
|
|
MD4_STEP_REV (MD4_H_S, a_rev, b_rev, c_rev, d_rev, 0, MD4C02, MD4S20);
|
|
|
|
const u32 sav_c = c_rev;
|
|
const u32 sav_d = d_rev;
|
|
|
|
MD4_STEP_REV1(MD4_G_S, b_rev, c_rev, d_rev, a_rev, w[15], MD4C01, MD4S13);
|
|
MD4_STEP_REV1(MD4_G_S, c_rev, d_rev, a_rev, b_rev, w[11], MD4C01, MD4S12);
|
|
|
|
/**
|
|
* loop
|
|
*/
|
|
|
|
u32 w0l = w[0];
|
|
|
|
for (u32 il_pos = 0; il_pos < IL_CNT; il_pos += VECT_SIZE)
|
|
{
|
|
const u32x w0r = words_buf_r[il_pos / VECT_SIZE];
|
|
|
|
const u32x w0 = w0l | w0r;
|
|
|
|
u32x pre_a = a_rev;
|
|
u32x pre_b = b_rev;
|
|
u32x pre_c = c_rev;
|
|
|
|
pre_a = pre_a - w0;
|
|
pre_b = pre_b - MD4_G (sav_c, sav_d, pre_a);
|
|
pre_c = pre_c - MD4_G (sav_d, pre_a, pre_b);
|
|
|
|
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, F_w0c00, MD4S00);
|
|
MD4_STEP0(MD4_Fo, d, a, b, c, F_w1c00, MD4S01);
|
|
MD4_STEP0(MD4_Fo, c, d, a, b, F_w2c00, MD4S02);
|
|
MD4_STEP0(MD4_Fo, b, c, d, a, F_w3c00, MD4S03);
|
|
MD4_STEP0(MD4_Fo, a, b, c, d, F_w4c00, MD4S00);
|
|
MD4_STEP0(MD4_Fo, d, a, b, c, F_w5c00, MD4S01);
|
|
MD4_STEP0(MD4_Fo, c, d, a, b, F_w6c00, MD4S02);
|
|
MD4_STEP0(MD4_Fo, b, c, d, a, F_w7c00, MD4S03);
|
|
MD4_STEP0(MD4_Fo, a, b, c, d, F_w8c00, MD4S00);
|
|
MD4_STEP0(MD4_Fo, d, a, b, c, F_w9c00, MD4S01);
|
|
MD4_STEP0(MD4_Fo, c, d, a, b, F_wac00, MD4S02);
|
|
MD4_STEP0(MD4_Fo, b, c, d, a, F_wbc00, MD4S03);
|
|
MD4_STEP0(MD4_Fo, a, b, c, d, F_wcc00, MD4S00);
|
|
MD4_STEP0(MD4_Fo, d, a, b, c, F_wdc00, MD4S01);
|
|
MD4_STEP0(MD4_Fo, c, d, a, b, F_wec00, MD4S02);
|
|
MD4_STEP0(MD4_Fo, b, c, d, a, F_wfc00, MD4S03);
|
|
|
|
MD4_STEP (MD4_Go, a, b, c, d, w0, G_w0c01, MD4S10);
|
|
MD4_STEP0(MD4_Go, d, a, b, c, G_w4c01, MD4S11);
|
|
MD4_STEP0(MD4_Go, c, d, a, b, G_w8c01, MD4S12);
|
|
MD4_STEP0(MD4_Go, b, c, d, a, G_wcc01, MD4S13);
|
|
MD4_STEP0(MD4_Go, a, b, c, d, G_w1c01, MD4S10);
|
|
MD4_STEP0(MD4_Go, d, a, b, c, G_w5c01, MD4S11);
|
|
MD4_STEP0(MD4_Go, c, d, a, b, G_w9c01, MD4S12);
|
|
MD4_STEP0(MD4_Go, b, c, d, a, G_wdc01, MD4S13);
|
|
MD4_STEP0(MD4_Go, a, b, c, d, G_w2c01, MD4S10);
|
|
MD4_STEP0(MD4_Go, d, a, b, c, G_w6c01, MD4S11);
|
|
MD4_STEP0(MD4_Go, c, d, a, b, G_wac01, MD4S12); if (MATCHES_NONE_VV (c, pre_c)) continue;
|
|
MD4_STEP0(MD4_Go, b, c, d, a, G_wec01, MD4S13); if (MATCHES_NONE_VV (b, pre_b)) continue;
|
|
MD4_STEP0(MD4_Go, a, b, c, d, G_w3c01, MD4S10); if (MATCHES_NONE_VV (a, pre_a)) continue;
|
|
MD4_STEP0(MD4_Go, d, a, b, c, G_w7c01, MD4S11);
|
|
MD4_STEP0(MD4_Go, c, d, a, b, G_wbc01, MD4S12);
|
|
MD4_STEP0(MD4_Go, b, c, d, a, G_wfc01, MD4S13);
|
|
|
|
MD4_STEP (MD4_H , a, b, c, d, w0, H_w0c02, MD4S20);
|
|
MD4_STEP0(MD4_H , d, a, b, c, H_w8c02, MD4S21);
|
|
MD4_STEP0(MD4_H , c, d, a, b, H_w4c02, MD4S22);
|
|
MD4_STEP0(MD4_H , b, c, d, a, H_wcc02, MD4S23);
|
|
MD4_STEP0(MD4_H , a, b, c, d, H_w2c02, MD4S20);
|
|
MD4_STEP0(MD4_H , d, a, b, c, H_wac02, MD4S21);
|
|
MD4_STEP0(MD4_H , c, d, a, b, H_w6c02, MD4S22);
|
|
MD4_STEP0(MD4_H , b, c, d, a, H_wec02, MD4S23);
|
|
MD4_STEP0(MD4_H , a, b, c, d, H_w1c02, MD4S20);
|
|
MD4_STEP0(MD4_H , d, a, b, c, H_w9c02, MD4S21);
|
|
MD4_STEP0(MD4_H , c, d, a, b, H_w5c02, MD4S22);
|
|
MD4_STEP0(MD4_H , b, c, d, a, H_wdc02, MD4S23);
|
|
MD4_STEP0(MD4_H , a, b, c, d, H_w3c02, MD4S20);
|
|
MD4_STEP0(MD4_H , d, a, b, c, H_wbc02, MD4S21);
|
|
MD4_STEP0(MD4_H , c, d, a, b, H_w7c02, MD4S22);
|
|
MD4_STEP0(MD4_H , b, c, d, a, H_wfc02, MD4S23);
|
|
|
|
COMPARE_S_SIMD (a, d, c, b);
|
|
}
|
|
}
|
|
|
|
KERNEL_FQ void m01000_m04 (KERN_ATTR_VECTOR ())
|
|
{
|
|
/**
|
|
* base
|
|
*/
|
|
|
|
const u64 lid = get_local_id (0);
|
|
const u64 gid = get_global_id (0);
|
|
const u64 lsz = get_local_size (0);
|
|
|
|
if (gid >= GID_CNT) return;
|
|
|
|
u32 w[16];
|
|
|
|
w[ 0] = pws[gid].i[ 0];
|
|
w[ 1] = pws[gid].i[ 1];
|
|
w[ 2] = pws[gid].i[ 2];
|
|
w[ 3] = pws[gid].i[ 3];
|
|
w[ 4] = 0;
|
|
w[ 5] = 0;
|
|
w[ 6] = 0;
|
|
w[ 7] = 0;
|
|
w[ 8] = 0;
|
|
w[ 9] = 0;
|
|
w[10] = 0;
|
|
w[11] = 0;
|
|
w[12] = 0;
|
|
w[13] = 0;
|
|
w[14] = pws[gid].i[14];
|
|
w[15] = 0;
|
|
|
|
const u32 pw_len = pws[gid].pw_len & 63;
|
|
|
|
/**
|
|
* main
|
|
*/
|
|
|
|
m01000m (w, pw_len, pws, rules_buf, combs_buf, words_buf_r, tmps, hooks, bitmaps_buf_s1_a, bitmaps_buf_s1_b, bitmaps_buf_s1_c, bitmaps_buf_s1_d, bitmaps_buf_s2_a, bitmaps_buf_s2_b, bitmaps_buf_s2_c, bitmaps_buf_s2_d, plains_buf, digests_buf, hashes_shown, salt_bufs, esalt_bufs, d_return_buf, d_extra0_buf, d_extra1_buf, d_extra2_buf, d_extra3_buf, kernel_param, gid, lid, lsz);
|
|
}
|
|
|
|
KERNEL_FQ void m01000_m08 (KERN_ATTR_VECTOR ())
|
|
{
|
|
/**
|
|
* base
|
|
*/
|
|
|
|
const u64 lid = get_local_id (0);
|
|
const u64 gid = get_global_id (0);
|
|
const u64 lsz = get_local_size (0);
|
|
|
|
if (gid >= GID_CNT) return;
|
|
|
|
u32 w[16];
|
|
|
|
w[ 0] = pws[gid].i[ 0];
|
|
w[ 1] = pws[gid].i[ 1];
|
|
w[ 2] = pws[gid].i[ 2];
|
|
w[ 3] = pws[gid].i[ 3];
|
|
w[ 4] = pws[gid].i[ 4];
|
|
w[ 5] = pws[gid].i[ 5];
|
|
w[ 6] = pws[gid].i[ 6];
|
|
w[ 7] = pws[gid].i[ 7];
|
|
w[ 8] = 0;
|
|
w[ 9] = 0;
|
|
w[10] = 0;
|
|
w[11] = 0;
|
|
w[12] = 0;
|
|
w[13] = 0;
|
|
w[14] = pws[gid].i[14];
|
|
w[15] = 0;
|
|
|
|
const u32 pw_len = pws[gid].pw_len & 63;
|
|
|
|
/**
|
|
* main
|
|
*/
|
|
|
|
m01000m (w, pw_len, pws, rules_buf, combs_buf, words_buf_r, tmps, hooks, bitmaps_buf_s1_a, bitmaps_buf_s1_b, bitmaps_buf_s1_c, bitmaps_buf_s1_d, bitmaps_buf_s2_a, bitmaps_buf_s2_b, bitmaps_buf_s2_c, bitmaps_buf_s2_d, plains_buf, digests_buf, hashes_shown, salt_bufs, esalt_bufs, d_return_buf, d_extra0_buf, d_extra1_buf, d_extra2_buf, d_extra3_buf, kernel_param, gid, lid, lsz);
|
|
}
|
|
|
|
KERNEL_FQ void m01000_m16 (KERN_ATTR_VECTOR ())
|
|
{
|
|
/**
|
|
* base
|
|
*/
|
|
|
|
const u64 lid = get_local_id (0);
|
|
const u64 gid = get_global_id (0);
|
|
const u64 lsz = get_local_size (0);
|
|
|
|
if (gid >= GID_CNT) return;
|
|
|
|
u32 w[16];
|
|
|
|
w[ 0] = pws[gid].i[ 0];
|
|
w[ 1] = pws[gid].i[ 1];
|
|
w[ 2] = pws[gid].i[ 2];
|
|
w[ 3] = pws[gid].i[ 3];
|
|
w[ 4] = pws[gid].i[ 4];
|
|
w[ 5] = pws[gid].i[ 5];
|
|
w[ 6] = pws[gid].i[ 6];
|
|
w[ 7] = pws[gid].i[ 7];
|
|
w[ 8] = pws[gid].i[ 8];
|
|
w[ 9] = pws[gid].i[ 9];
|
|
w[10] = pws[gid].i[10];
|
|
w[11] = pws[gid].i[11];
|
|
w[12] = pws[gid].i[12];
|
|
w[13] = pws[gid].i[13];
|
|
w[14] = pws[gid].i[14];
|
|
w[15] = pws[gid].i[15];
|
|
|
|
const u32 pw_len = pws[gid].pw_len & 63;
|
|
|
|
/**
|
|
* main
|
|
*/
|
|
|
|
m01000m (w, pw_len, pws, rules_buf, combs_buf, words_buf_r, tmps, hooks, bitmaps_buf_s1_a, bitmaps_buf_s1_b, bitmaps_buf_s1_c, bitmaps_buf_s1_d, bitmaps_buf_s2_a, bitmaps_buf_s2_b, bitmaps_buf_s2_c, bitmaps_buf_s2_d, plains_buf, digests_buf, hashes_shown, salt_bufs, esalt_bufs, d_return_buf, d_extra0_buf, d_extra1_buf, d_extra2_buf, d_extra3_buf, kernel_param, gid, lid, lsz);
|
|
}
|
|
|
|
KERNEL_FQ void m01000_s04 (KERN_ATTR_VECTOR ())
|
|
{
|
|
/**
|
|
* base
|
|
*/
|
|
|
|
const u64 lid = get_local_id (0);
|
|
const u64 gid = get_global_id (0);
|
|
const u64 lsz = get_local_size (0);
|
|
|
|
if (gid >= GID_CNT) return;
|
|
|
|
u32 w[16];
|
|
|
|
w[ 0] = pws[gid].i[ 0];
|
|
w[ 1] = pws[gid].i[ 1];
|
|
w[ 2] = pws[gid].i[ 2];
|
|
w[ 3] = pws[gid].i[ 3];
|
|
w[ 4] = 0;
|
|
w[ 5] = 0;
|
|
w[ 6] = 0;
|
|
w[ 7] = 0;
|
|
w[ 8] = 0;
|
|
w[ 9] = 0;
|
|
w[10] = 0;
|
|
w[11] = 0;
|
|
w[12] = 0;
|
|
w[13] = 0;
|
|
w[14] = pws[gid].i[14];
|
|
w[15] = 0;
|
|
|
|
const u32 pw_len = pws[gid].pw_len & 63;
|
|
|
|
/**
|
|
* main
|
|
*/
|
|
|
|
m01000s (w, pw_len, pws, rules_buf, combs_buf, words_buf_r, tmps, hooks, bitmaps_buf_s1_a, bitmaps_buf_s1_b, bitmaps_buf_s1_c, bitmaps_buf_s1_d, bitmaps_buf_s2_a, bitmaps_buf_s2_b, bitmaps_buf_s2_c, bitmaps_buf_s2_d, plains_buf, digests_buf, hashes_shown, salt_bufs, esalt_bufs, d_return_buf, d_extra0_buf, d_extra1_buf, d_extra2_buf, d_extra3_buf, kernel_param, gid, lid, lsz);
|
|
}
|
|
|
|
KERNEL_FQ void m01000_s08 (KERN_ATTR_VECTOR ())
|
|
{
|
|
/**
|
|
* base
|
|
*/
|
|
|
|
const u64 lid = get_local_id (0);
|
|
const u64 gid = get_global_id (0);
|
|
const u64 lsz = get_local_size (0);
|
|
|
|
if (gid >= GID_CNT) return;
|
|
|
|
u32 w[16];
|
|
|
|
w[ 0] = pws[gid].i[ 0];
|
|
w[ 1] = pws[gid].i[ 1];
|
|
w[ 2] = pws[gid].i[ 2];
|
|
w[ 3] = pws[gid].i[ 3];
|
|
w[ 4] = pws[gid].i[ 4];
|
|
w[ 5] = pws[gid].i[ 5];
|
|
w[ 6] = pws[gid].i[ 6];
|
|
w[ 7] = pws[gid].i[ 7];
|
|
w[ 8] = 0;
|
|
w[ 9] = 0;
|
|
w[10] = 0;
|
|
w[11] = 0;
|
|
w[12] = 0;
|
|
w[13] = 0;
|
|
w[14] = pws[gid].i[14];
|
|
w[15] = 0;
|
|
|
|
const u32 pw_len = pws[gid].pw_len & 63;
|
|
|
|
/**
|
|
* main
|
|
*/
|
|
|
|
m01000s (w, pw_len, pws, rules_buf, combs_buf, words_buf_r, tmps, hooks, bitmaps_buf_s1_a, bitmaps_buf_s1_b, bitmaps_buf_s1_c, bitmaps_buf_s1_d, bitmaps_buf_s2_a, bitmaps_buf_s2_b, bitmaps_buf_s2_c, bitmaps_buf_s2_d, plains_buf, digests_buf, hashes_shown, salt_bufs, esalt_bufs, d_return_buf, d_extra0_buf, d_extra1_buf, d_extra2_buf, d_extra3_buf, kernel_param, gid, lid, lsz);
|
|
}
|
|
|
|
KERNEL_FQ void m01000_s16 (KERN_ATTR_VECTOR ())
|
|
{
|
|
/**
|
|
* base
|
|
*/
|
|
|
|
const u64 lid = get_local_id (0);
|
|
const u64 gid = get_global_id (0);
|
|
const u64 lsz = get_local_size (0);
|
|
|
|
if (gid >= GID_CNT) return;
|
|
|
|
u32 w[16];
|
|
|
|
w[ 0] = pws[gid].i[ 0];
|
|
w[ 1] = pws[gid].i[ 1];
|
|
w[ 2] = pws[gid].i[ 2];
|
|
w[ 3] = pws[gid].i[ 3];
|
|
w[ 4] = pws[gid].i[ 4];
|
|
w[ 5] = pws[gid].i[ 5];
|
|
w[ 6] = pws[gid].i[ 6];
|
|
w[ 7] = pws[gid].i[ 7];
|
|
w[ 8] = pws[gid].i[ 8];
|
|
w[ 9] = pws[gid].i[ 9];
|
|
w[10] = pws[gid].i[10];
|
|
w[11] = pws[gid].i[11];
|
|
w[12] = pws[gid].i[12];
|
|
w[13] = pws[gid].i[13];
|
|
w[14] = pws[gid].i[14];
|
|
w[15] = pws[gid].i[15];
|
|
|
|
const u32 pw_len = pws[gid].pw_len & 63;
|
|
|
|
/**
|
|
* main
|
|
*/
|
|
|
|
m01000s (w, pw_len, pws, rules_buf, combs_buf, words_buf_r, tmps, hooks, bitmaps_buf_s1_a, bitmaps_buf_s1_b, bitmaps_buf_s1_c, bitmaps_buf_s1_d, bitmaps_buf_s2_a, bitmaps_buf_s2_b, bitmaps_buf_s2_c, bitmaps_buf_s2_d, plains_buf, digests_buf, hashes_shown, salt_bufs, esalt_bufs, d_return_buf, d_extra0_buf, d_extra1_buf, d_extra2_buf, d_extra3_buf, kernel_param, gid, lid, lsz);
|
|
}
|