1
0
mirror of https://github.com/hashcat/hashcat.git synced 2024-10-31 20:48:57 +00:00
hashcat/OpenCL/m00600_a0-optimized.cl

202 lines
4.1 KiB
Common Lisp
Raw Permalink Normal View History

2017-04-22 13:51:39 +00:00
/**
* Author......: See docs/credits.txt
* License.....: MIT
*/
#define NEW_SIMD_CODE
#ifdef KERNEL_STATIC
#include "inc_vendor.h"
#include "inc_types.h"
2019-04-26 11:59:43 +00:00
#include "inc_platform.cl"
2017-04-22 13:51:39 +00:00
#include "inc_common.cl"
2017-08-11 09:25:47 +00:00
#include "inc_rp_optimized.h"
#include "inc_rp_optimized.cl"
2017-04-22 13:51:39 +00:00
#include "inc_simd.cl"
#include "inc_hash_blake2b.cl"
#endif
2017-04-22 13:51:39 +00:00
KERNEL_FQ void m00600_m04 (KERN_ATTR_RULES ())
2017-10-09 16:45:11 +00:00
{
2017-04-22 13:51:39 +00:00
/**
* base
2017-04-22 13:51:39 +00:00
*/
const u64 gid = get_global_id (0);
if (gid >= gid_max) return;
2017-04-22 13:51:39 +00:00
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;
2017-04-22 13:51:39 +00:00
/**
* loop
*/
for (u32 il_pos = 0; il_pos < il_cnt; il_pos += VECT_SIZE)
{
2017-04-22 13:51:39 +00:00
u32x w0[4] = { 0 };
u32x w1[4] = { 0 };
u32x w2[4] = { 0 };
u32x w3[4] = { 0 };
const u32x out_len = apply_rules_vect_optimized (pw_buf0, pw_buf1, pw_len, rules_buf, il_pos, w0, w1);
2017-04-22 13:51:39 +00:00
u64x m[16];
m[ 0] = hl32_to_64 (w0[1], w0[0]);
m[ 1] = hl32_to_64 (w0[3], w0[2]);
m[ 2] = hl32_to_64 (w1[1], w1[0]);
m[ 3] = hl32_to_64 (w1[3], w1[2]);
m[ 4] = hl32_to_64 (w2[1], w2[0]);
m[ 5] = hl32_to_64 (w2[3], w2[2]);
m[ 6] = hl32_to_64 (w3[1], w3[0]);
m[ 7] = hl32_to_64 (w3[3], w3[2]);
m[ 8] = 0;
m[ 9] = 0;
m[10] = 0;
m[11] = 0;
m[12] = 0;
m[13] = 0;
m[14] = 0;
m[15] = 0;
u64x h[8];
h[0] = BLAKE2B_IV_00 ^ 0x01010040;
h[1] = BLAKE2B_IV_01;
h[2] = BLAKE2B_IV_02;
h[3] = BLAKE2B_IV_03;
h[4] = BLAKE2B_IV_04;
h[5] = BLAKE2B_IV_05;
h[6] = BLAKE2B_IV_06;
h[7] = BLAKE2B_IV_07;
blake2b_transform_vector (h, m, out_len, BLAKE2B_FINAL);
const u32x r0 = h32_from_64 (h[0]);
const u32x r1 = l32_from_64 (h[0]);
const u32x r2 = h32_from_64 (h[1]);
const u32x r3 = l32_from_64 (h[1]);
2017-04-22 13:51:39 +00:00
2020-06-13 09:19:00 +00:00
COMPARE_M_SIMD (r0, r1, r2, r3);
2017-10-09 16:45:11 +00:00
}
2017-04-23 01:50:22 +00:00
}
KERNEL_FQ void m00600_m08 (KERN_ATTR_RULES ())
2017-04-23 01:50:22 +00:00
{
2017-04-22 13:51:39 +00:00
}
KERNEL_FQ void m00600_m16 (KERN_ATTR_RULES ())
2017-04-22 13:51:39 +00:00
{
2017-04-23 01:50:22 +00:00
}
KERNEL_FQ void m00600_s04 (KERN_ATTR_RULES ())
2017-10-09 16:45:11 +00:00
{
2017-04-22 13:51:39 +00:00
/**
* base
2017-04-22 13:51:39 +00:00
*/
const u64 gid = get_global_id (0);
2017-04-22 13:51:39 +00:00
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;
2017-04-22 13:51:39 +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]
};
2017-10-09 16:45:11 +00:00
2017-04-22 13:51:39 +00:00
/**
* loop
*/
for (u32 il_pos = 0; il_pos < il_cnt; il_pos += VECT_SIZE)
{
2017-04-22 13:51:39 +00:00
u32x w0[4] = { 0 };
u32x w1[4] = { 0 };
u32x w2[4] = { 0 };
u32x w3[4] = { 0 };
2017-10-09 16:45:11 +00:00
const u32x out_len = apply_rules_vect_optimized (pw_buf0, pw_buf1, pw_len, rules_buf, il_pos, w0, w1);
2017-04-22 13:51:39 +00:00
u64x m[16];
m[ 0] = hl32_to_64 (w0[1], w0[0]);
m[ 1] = hl32_to_64 (w0[3], w0[2]);
m[ 2] = hl32_to_64 (w1[1], w1[0]);
m[ 3] = hl32_to_64 (w1[3], w1[2]);
m[ 4] = hl32_to_64 (w2[1], w2[0]);
m[ 5] = hl32_to_64 (w2[3], w2[2]);
m[ 6] = hl32_to_64 (w3[1], w3[0]);
m[ 7] = hl32_to_64 (w3[3], w3[2]);
m[ 8] = 0;
m[ 9] = 0;
m[10] = 0;
m[11] = 0;
m[12] = 0;
m[13] = 0;
m[14] = 0;
m[15] = 0;
u64x h[8];
h[0] = BLAKE2B_IV_00 ^ 0x01010040;
h[1] = BLAKE2B_IV_01;
h[2] = BLAKE2B_IV_02;
h[3] = BLAKE2B_IV_03;
h[4] = BLAKE2B_IV_04;
h[5] = BLAKE2B_IV_05;
h[6] = BLAKE2B_IV_06;
h[7] = BLAKE2B_IV_07;
blake2b_transform_vector (h, m, out_len, BLAKE2B_FINAL);
const u32x r0 = h32_from_64 (h[0]);
const u32x r1 = l32_from_64 (h[0]);
const u32x r2 = h32_from_64 (h[1]);
const u32x r3 = l32_from_64 (h[1]);
2017-04-22 13:51:39 +00:00
2020-06-13 09:19:00 +00:00
COMPARE_S_SIMD (r0, r1, r2, r3);
2017-10-09 16:45:11 +00:00
}
2017-04-22 13:51:39 +00:00
}
KERNEL_FQ void m00600_s08 (KERN_ATTR_RULES ())
2017-04-22 13:51:39 +00:00
{
}
KERNEL_FQ void m00600_s16 (KERN_ATTR_RULES ())
2017-04-22 13:51:39 +00:00
{
}