mirror of
https://github.com/hashcat/hashcat.git
synced 2024-11-05 06:58:56 +00:00
184 lines
3.3 KiB
Common Lisp
184 lines
3.3 KiB
Common Lisp
/**
|
|
* Author......: See docs/credits.txt
|
|
* License.....: MIT
|
|
*/
|
|
|
|
//#define NEW_SIMD_CODE
|
|
|
|
#ifdef KERNEL_STATIC
|
|
#include "inc_vendor.h"
|
|
#include "inc_types.h"
|
|
#include "inc_common.cl"
|
|
#include "inc_rp_optimized.h"
|
|
#include "inc_rp_optimized.cl"
|
|
#include "inc_simd.cl"
|
|
#endif
|
|
|
|
DECLSPEC u32 hashCode (const u32 init, const u32 *w, const u32 pw_len)
|
|
{
|
|
u32 hash = init;
|
|
|
|
for (u32 i = 0; i < pw_len; i += 4)
|
|
{
|
|
u32 tmp = w[i / 4];
|
|
|
|
const u32 left = pw_len - i;
|
|
|
|
const u32 c = (left > 4) ? 4 : left;
|
|
|
|
switch (c)
|
|
{
|
|
case 1:
|
|
hash *= 31; hash += (tmp >> 0) & 0xff;
|
|
break;
|
|
case 2:
|
|
hash *= 31; hash += (tmp >> 0) & 0xff;
|
|
hash *= 31; hash += (tmp >> 8) & 0xff;
|
|
break;
|
|
case 3:
|
|
hash *= 31; hash += (tmp >> 0) & 0xff;
|
|
hash *= 31; hash += (tmp >> 8) & 0xff;
|
|
hash *= 31; hash += (tmp >> 16) & 0xff;
|
|
break;
|
|
case 4:
|
|
hash *= 31; hash += (tmp >> 0) & 0xff;
|
|
hash *= 31; hash += (tmp >> 8) & 0xff;
|
|
hash *= 31; hash += (tmp >> 16) & 0xff;
|
|
hash *= 31; hash += (tmp >> 24) & 0xff;
|
|
break;
|
|
}
|
|
}
|
|
|
|
return hash;
|
|
}
|
|
|
|
KERNEL_FQ void m18700_m04 (KERN_ATTR_RULES ())
|
|
{
|
|
/**
|
|
* modifier
|
|
*/
|
|
|
|
const u64 lid = get_local_id (0);
|
|
|
|
/**
|
|
* base
|
|
*/
|
|
|
|
const u64 gid = get_global_id (0);
|
|
|
|
if (gid >= gid_max) return;
|
|
|
|
u32 pw_buf0[4];
|
|
u32 pw_buf1[4];
|
|
|
|
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];
|
|
|
|
const u32 pw_len = pws[gid].pw_len & 63;
|
|
|
|
/**
|
|
* loop
|
|
*/
|
|
|
|
for (u32 il_pos = 0; il_pos < il_cnt; il_pos += VECT_SIZE)
|
|
{
|
|
u32x w[16] = { 0 };
|
|
|
|
const u32x out_len = apply_rules_vect_optimized (pw_buf0, pw_buf1, pw_len, rules_buf, il_pos, w + 0, w + 4);
|
|
|
|
u32x hash = hashCode (0, w, out_len);
|
|
|
|
const u32x r0 = hash;
|
|
const u32x r1 = 0;
|
|
const u32x r2 = 0;
|
|
const u32x r3 = 0;
|
|
|
|
COMPARE_M_SIMD (r0, r1, r2, r3);
|
|
}
|
|
}
|
|
|
|
KERNEL_FQ void m18700_m08 (KERN_ATTR_RULES ())
|
|
{
|
|
}
|
|
|
|
KERNEL_FQ void m18700_m16 (KERN_ATTR_RULES ())
|
|
{
|
|
}
|
|
|
|
KERNEL_FQ void m18700_s04 (KERN_ATTR_RULES ())
|
|
{
|
|
/**
|
|
* modifier
|
|
*/
|
|
|
|
const u64 lid = get_local_id (0);
|
|
|
|
/**
|
|
* base
|
|
*/
|
|
|
|
const u64 gid = get_global_id (0);
|
|
|
|
if (gid >= gid_max) return;
|
|
|
|
u32 pw_buf0[4];
|
|
u32 pw_buf1[4];
|
|
|
|
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];
|
|
|
|
const u32 pw_len = pws[gid].pw_len & 63;
|
|
|
|
/**
|
|
* digest
|
|
*/
|
|
|
|
const u32 search[4] =
|
|
{
|
|
digests_buf[digests_offset].digest_buf[DGST_R0],
|
|
0,
|
|
0,
|
|
0
|
|
};
|
|
|
|
/**
|
|
* loop
|
|
*/
|
|
|
|
for (u32 il_pos = 0; il_pos < il_cnt; il_pos += VECT_SIZE)
|
|
{
|
|
u32x w[16] = { 0 };
|
|
|
|
const u32x out_len = apply_rules_vect_optimized (pw_buf0, pw_buf1, pw_len, rules_buf, il_pos, w + 0, w + 4);
|
|
|
|
u32x hash = hashCode (0, w, out_len);
|
|
|
|
const u32x r0 = hash;
|
|
const u32x r1 = 0;
|
|
const u32x r2 = 0;
|
|
const u32x r3 = 0;
|
|
|
|
COMPARE_S_SIMD (r0, r1, r2, r3);
|
|
}
|
|
}
|
|
|
|
KERNEL_FQ void m18700_s08 (KERN_ATTR_RULES ())
|
|
{
|
|
}
|
|
|
|
KERNEL_FQ void m18700_s16 (KERN_ATTR_RULES ())
|
|
{
|
|
}
|