mirror of
https://github.com/hashcat/hashcat.git
synced 2024-11-14 03:39:09 +00:00
0bf4e3c34a
- Some performance on low-end GPU may drop because of that, but only for a few hash-modes - Dropped scalar code (aka warp) since we do not have any vector datatypes anymore - Renamed C++ overloading functions memcat32_9 -> memcat_c32_w4x4_a3x4 - Still need to fix kernels to new function names, needs to be done manually - Temperature Management needs to be rewritten partially because of conflicting datatypes names - Added code to create different codepaths for NV on AMD in runtime in host (see data.vendor_id) - Added code to create different codepaths for NV on AMD in runtime in kernels (see IS_NV and IS_AMD) - First tests working for -m 0, for example - Great performance increases in general for NV so far - Tested amp_* and markov_* kernel - Migrated special NV optimizations for rule processor
350 lines
11 KiB
Common Lisp
350 lines
11 KiB
Common Lisp
/**
|
|
* Author......: Jens Steube <jens.steube@gmail.com>
|
|
* License.....: MIT
|
|
*/
|
|
|
|
#define _MD5_
|
|
|
|
#include "include/constants.h"
|
|
#include "include/kernel_vendor.h"
|
|
|
|
#define DGST_R0 0
|
|
#define DGST_R1 1
|
|
#define DGST_R2 2
|
|
#define DGST_R3 3
|
|
|
|
#include "include/kernel_functions.c"
|
|
#include "types_ocl.c"
|
|
#include "common.c"
|
|
|
|
#define COMPARE_S "check_single_comp4.c"
|
|
#define COMPARE_M "check_multi_comp4.c"
|
|
|
|
static void md5_transform (const u32 w0[4], const u32 w1[4], const u32 w2[4], const u32 w3[4], u32 digest[4])
|
|
{
|
|
u32 a = digest[0];
|
|
u32 b = digest[1];
|
|
u32 c = digest[2];
|
|
u32 d = digest[3];
|
|
|
|
u32 w0_t = w0[0];
|
|
u32 w1_t = w0[1];
|
|
u32 w2_t = w0[2];
|
|
u32 w3_t = w0[3];
|
|
u32 w4_t = w1[0];
|
|
u32 w5_t = w1[1];
|
|
u32 w6_t = w1[2];
|
|
u32 w7_t = w1[3];
|
|
u32 w8_t = w2[0];
|
|
u32 w9_t = w2[1];
|
|
u32 wa_t = w2[2];
|
|
u32 wb_t = w2[3];
|
|
u32 wc_t = w3[0];
|
|
u32 wd_t = w3[1];
|
|
u32 we_t = w3[2];
|
|
u32 wf_t = 0;
|
|
|
|
u32 tmp2;
|
|
|
|
MD5_STEP (MD5_Fo, a, b, c, d, w0_t, MD5C00, MD5S00);
|
|
MD5_STEP (MD5_Fo, d, a, b, c, w1_t, MD5C01, MD5S01);
|
|
MD5_STEP (MD5_Fo, c, d, a, b, w2_t, MD5C02, MD5S02);
|
|
MD5_STEP (MD5_Fo, b, c, d, a, w3_t, MD5C03, MD5S03);
|
|
MD5_STEP (MD5_Fo, a, b, c, d, w4_t, MD5C04, MD5S00);
|
|
MD5_STEP (MD5_Fo, d, a, b, c, w5_t, MD5C05, MD5S01);
|
|
MD5_STEP (MD5_Fo, c, d, a, b, w6_t, MD5C06, MD5S02);
|
|
MD5_STEP (MD5_Fo, b, c, d, a, w7_t, MD5C07, MD5S03);
|
|
MD5_STEP (MD5_Fo, a, b, c, d, w8_t, MD5C08, MD5S00);
|
|
MD5_STEP (MD5_Fo, d, a, b, c, w9_t, MD5C09, MD5S01);
|
|
MD5_STEP (MD5_Fo, c, d, a, b, wa_t, MD5C0a, MD5S02);
|
|
MD5_STEP (MD5_Fo, b, c, d, a, wb_t, MD5C0b, MD5S03);
|
|
MD5_STEP (MD5_Fo, a, b, c, d, wc_t, MD5C0c, MD5S00);
|
|
MD5_STEP (MD5_Fo, d, a, b, c, wd_t, MD5C0d, MD5S01);
|
|
MD5_STEP (MD5_Fo, c, d, a, b, we_t, MD5C0e, MD5S02);
|
|
MD5_STEP (MD5_Fo, b, c, d, a, wf_t, MD5C0f, MD5S03);
|
|
|
|
MD5_STEP (MD5_Go, a, b, c, d, w1_t, MD5C10, MD5S10);
|
|
MD5_STEP (MD5_Go, d, a, b, c, w6_t, MD5C11, MD5S11);
|
|
MD5_STEP (MD5_Go, c, d, a, b, wb_t, MD5C12, MD5S12);
|
|
MD5_STEP (MD5_Go, b, c, d, a, w0_t, MD5C13, MD5S13);
|
|
MD5_STEP (MD5_Go, a, b, c, d, w5_t, MD5C14, MD5S10);
|
|
MD5_STEP (MD5_Go, d, a, b, c, wa_t, MD5C15, MD5S11);
|
|
MD5_STEP (MD5_Go, c, d, a, b, wf_t, MD5C16, MD5S12);
|
|
MD5_STEP (MD5_Go, b, c, d, a, w4_t, MD5C17, MD5S13);
|
|
MD5_STEP (MD5_Go, a, b, c, d, w9_t, MD5C18, MD5S10);
|
|
MD5_STEP (MD5_Go, d, a, b, c, we_t, MD5C19, MD5S11);
|
|
MD5_STEP (MD5_Go, c, d, a, b, w3_t, MD5C1a, MD5S12);
|
|
MD5_STEP (MD5_Go, b, c, d, a, w8_t, MD5C1b, MD5S13);
|
|
MD5_STEP (MD5_Go, a, b, c, d, wd_t, MD5C1c, MD5S10);
|
|
MD5_STEP (MD5_Go, d, a, b, c, w2_t, MD5C1d, MD5S11);
|
|
MD5_STEP (MD5_Go, c, d, a, b, w7_t, MD5C1e, MD5S12);
|
|
MD5_STEP (MD5_Go, b, c, d, a, wc_t, MD5C1f, MD5S13);
|
|
|
|
MD5_STEP (MD5_H1, a, b, c, d, w5_t, MD5C20, MD5S20);
|
|
MD5_STEP (MD5_H2, d, a, b, c, w8_t, MD5C21, MD5S21);
|
|
MD5_STEP (MD5_H1, c, d, a, b, wb_t, MD5C22, MD5S22);
|
|
MD5_STEP (MD5_H2, b, c, d, a, we_t, MD5C23, MD5S23);
|
|
MD5_STEP (MD5_H1, a, b, c, d, w1_t, MD5C24, MD5S20);
|
|
MD5_STEP (MD5_H2, d, a, b, c, w4_t, MD5C25, MD5S21);
|
|
MD5_STEP (MD5_H1, c, d, a, b, w7_t, MD5C26, MD5S22);
|
|
MD5_STEP (MD5_H2, b, c, d, a, wa_t, MD5C27, MD5S23);
|
|
MD5_STEP (MD5_H1, a, b, c, d, wd_t, MD5C28, MD5S20);
|
|
MD5_STEP (MD5_H2, d, a, b, c, w0_t, MD5C29, MD5S21);
|
|
MD5_STEP (MD5_H1, c, d, a, b, w3_t, MD5C2a, MD5S22);
|
|
MD5_STEP (MD5_H2, b, c, d, a, w6_t, MD5C2b, MD5S23);
|
|
MD5_STEP (MD5_H1, a, b, c, d, w9_t, MD5C2c, MD5S20);
|
|
MD5_STEP (MD5_H2, d, a, b, c, wc_t, MD5C2d, MD5S21);
|
|
MD5_STEP (MD5_H1, c, d, a, b, wf_t, MD5C2e, MD5S22);
|
|
MD5_STEP (MD5_H2, b, c, d, a, w2_t, MD5C2f, MD5S23);
|
|
|
|
MD5_STEP (MD5_I , a, b, c, d, w0_t, MD5C30, MD5S30);
|
|
MD5_STEP (MD5_I , d, a, b, c, w7_t, MD5C31, MD5S31);
|
|
MD5_STEP (MD5_I , c, d, a, b, we_t, MD5C32, MD5S32);
|
|
MD5_STEP (MD5_I , b, c, d, a, w5_t, MD5C33, MD5S33);
|
|
MD5_STEP (MD5_I , a, b, c, d, wc_t, MD5C34, MD5S30);
|
|
MD5_STEP (MD5_I , d, a, b, c, w3_t, MD5C35, MD5S31);
|
|
MD5_STEP (MD5_I , c, d, a, b, wa_t, MD5C36, MD5S32);
|
|
MD5_STEP (MD5_I , b, c, d, a, w1_t, MD5C37, MD5S33);
|
|
MD5_STEP (MD5_I , a, b, c, d, w8_t, MD5C38, MD5S30);
|
|
MD5_STEP (MD5_I , d, a, b, c, wf_t, MD5C39, MD5S31);
|
|
MD5_STEP (MD5_I , c, d, a, b, w6_t, MD5C3a, MD5S32);
|
|
MD5_STEP (MD5_I , b, c, d, a, wd_t, MD5C3b, MD5S33);
|
|
MD5_STEP (MD5_I , a, b, c, d, w4_t, MD5C3c, MD5S30);
|
|
MD5_STEP (MD5_I , d, a, b, c, wb_t, MD5C3d, MD5S31);
|
|
MD5_STEP (MD5_I , c, d, a, b, w2_t, MD5C3e, MD5S32);
|
|
MD5_STEP (MD5_I , b, c, d, a, w9_t, MD5C3f, MD5S33);
|
|
|
|
digest[0] += a;
|
|
digest[1] += b;
|
|
digest[2] += c;
|
|
digest[3] += d;
|
|
}
|
|
|
|
__kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m00400_init (__global pw_t *pws, __global gpu_rule_t *rules_buf, __global comb_t *combs_buf, __global bf_t *bfs_buf, __global phpass_tmp_t *tmps, __global void *hooks, __global u32 *bitmaps_buf_s1_a, __global u32 *bitmaps_buf_s1_b, __global u32 *bitmaps_buf_s1_c, __global u32 *bitmaps_buf_s1_d, __global u32 *bitmaps_buf_s2_a, __global u32 *bitmaps_buf_s2_b, __global u32 *bitmaps_buf_s2_c, __global u32 *bitmaps_buf_s2_d, __global plain_t *plains_buf, __global digest_t *digests_buf, __global u32 *hashes_shown, __global salt_t *salt_bufs, __global void *esalt_bufs, __global u32 *d_return_buf, __global u32 *d_scryptV_buf, const u32 bitmap_mask, const u32 bitmap_shift1, const u32 bitmap_shift2, const u32 salt_pos, const u32 loop_pos, const u32 loop_cnt, const u32 rules_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
|
|
{
|
|
/**
|
|
* base
|
|
*/
|
|
|
|
const u32 gid = get_global_id (0);
|
|
|
|
if (gid >= gid_max) return;
|
|
|
|
u32 w0[4];
|
|
|
|
w0[0] = pws[gid].i[ 0];
|
|
w0[1] = pws[gid].i[ 1];
|
|
w0[2] = pws[gid].i[ 2];
|
|
w0[3] = pws[gid].i[ 3];
|
|
|
|
u32 w1[4];
|
|
|
|
w1[0] = pws[gid].i[ 4];
|
|
w1[1] = pws[gid].i[ 5];
|
|
w1[2] = pws[gid].i[ 6];
|
|
w1[3] = pws[gid].i[ 7];
|
|
|
|
u32 w2[4];
|
|
|
|
w2[0] = pws[gid].i[ 8];
|
|
w2[1] = pws[gid].i[ 9];
|
|
w2[2] = 0;
|
|
w2[3] = 0;
|
|
|
|
const u32 pw_len = pws[gid].pw_len;
|
|
|
|
/**
|
|
* salt
|
|
*/
|
|
|
|
u32 salt_buf[2];
|
|
|
|
salt_buf[0] = salt_bufs[salt_pos].salt_buf[0];
|
|
salt_buf[1] = salt_bufs[salt_pos].salt_buf[1];
|
|
|
|
/**
|
|
* init
|
|
*/
|
|
|
|
u32 block_len = 8 + pw_len;
|
|
|
|
u32 block0[4];
|
|
|
|
block0[0] = salt_buf[0];
|
|
block0[1] = salt_buf[1];
|
|
block0[2] = w0[0];
|
|
block0[3] = w0[1];
|
|
|
|
u32 block1[4];
|
|
|
|
block1[0] = w0[2];
|
|
block1[1] = w0[3];
|
|
block1[2] = w1[0];
|
|
block1[3] = w1[1];
|
|
|
|
u32 block2[4];
|
|
|
|
block2[0] = w1[2];
|
|
block2[1] = w1[3];
|
|
block2[2] = w2[0];
|
|
block2[3] = w2[1];
|
|
|
|
u32 block3[4];
|
|
|
|
block3[0] = 0;
|
|
block3[1] = 0;
|
|
block3[2] = block_len * 8;
|
|
block3[3] = 0;
|
|
|
|
append_0x80_4x4 (block0, block1, block2, block3, block_len);
|
|
|
|
/**
|
|
* init
|
|
*/
|
|
|
|
u32 digest[4];
|
|
|
|
digest[0] = MD5M_A;
|
|
digest[1] = MD5M_B;
|
|
digest[2] = MD5M_C;
|
|
digest[3] = MD5M_D;
|
|
|
|
md5_transform (block0, block1, block2, block3, digest);
|
|
|
|
tmps[gid].digest_buf[0] = digest[0];
|
|
tmps[gid].digest_buf[1] = digest[1];
|
|
tmps[gid].digest_buf[2] = digest[2];
|
|
tmps[gid].digest_buf[3] = digest[3];
|
|
}
|
|
|
|
__kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m00400_loop (__global pw_t *pws, __global gpu_rule_t *rules_buf, __global comb_t *combs_buf, __global bf_t *bfs_buf, __global phpass_tmp_t *tmps, __global void *hooks, __global u32 *bitmaps_buf_s1_a, __global u32 *bitmaps_buf_s1_b, __global u32 *bitmaps_buf_s1_c, __global u32 *bitmaps_buf_s1_d, __global u32 *bitmaps_buf_s2_a, __global u32 *bitmaps_buf_s2_b, __global u32 *bitmaps_buf_s2_c, __global u32 *bitmaps_buf_s2_d, __global plain_t *plains_buf, __global digest_t *digests_buf, __global u32 *hashes_shown, __global salt_t *salt_bufs, __global void *esalt_bufs, __global u32 *d_return_buf, __global u32 *d_scryptV_buf, const u32 bitmap_mask, const u32 bitmap_shift1, const u32 bitmap_shift2, const u32 salt_pos, const u32 loop_pos, const u32 loop_cnt, const u32 rules_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
|
|
{
|
|
/**
|
|
* base
|
|
*/
|
|
|
|
const u32 gid = get_global_id (0);
|
|
|
|
if (gid >= gid_max) return;
|
|
|
|
u32 w0[4];
|
|
|
|
w0[0] = pws[gid].i[ 0];
|
|
w0[1] = pws[gid].i[ 1];
|
|
w0[2] = pws[gid].i[ 2];
|
|
w0[3] = pws[gid].i[ 3];
|
|
|
|
u32 w1[4];
|
|
|
|
w1[0] = pws[gid].i[ 4];
|
|
w1[1] = pws[gid].i[ 5];
|
|
w1[2] = pws[gid].i[ 6];
|
|
w1[3] = pws[gid].i[ 7];
|
|
|
|
u32 w2[4];
|
|
|
|
w2[0] = pws[gid].i[ 8];
|
|
w2[1] = pws[gid].i[ 9];
|
|
w2[2] = 0;
|
|
w2[3] = 0;
|
|
|
|
const u32 pw_len = pws[gid].pw_len;
|
|
|
|
/**
|
|
* digest
|
|
*/
|
|
|
|
u32 digest[4];
|
|
|
|
digest[0] = tmps[gid].digest_buf[0];
|
|
digest[1] = tmps[gid].digest_buf[1];
|
|
digest[2] = tmps[gid].digest_buf[2];
|
|
digest[3] = tmps[gid].digest_buf[3];
|
|
|
|
/**
|
|
* loop
|
|
*/
|
|
|
|
u32 block_len = (16 + pw_len);
|
|
|
|
u32 block0[4];
|
|
|
|
block0[0] = 0;
|
|
block0[1] = 0;
|
|
block0[2] = 0;
|
|
block0[3] = 0;
|
|
|
|
u32 block1[4];
|
|
|
|
block1[0] = w0[0];
|
|
block1[1] = w0[1];
|
|
block1[2] = w0[2];
|
|
block1[3] = w0[3];
|
|
|
|
u32 block2[4];
|
|
|
|
block2[0] = w1[0];
|
|
block2[1] = w1[1];
|
|
block2[2] = w1[2];
|
|
block2[3] = w1[3];
|
|
|
|
u32 block3[4];
|
|
|
|
block3[0] = w2[0];
|
|
block3[1] = w2[1];
|
|
block3[2] = block_len * 8;
|
|
block3[3] = 0;
|
|
|
|
append_0x80_4x4 (block0, block1, block2, block3, block_len);
|
|
|
|
/**
|
|
* init
|
|
*/
|
|
|
|
for (u32 i = 0; i < loop_cnt; i++)
|
|
{
|
|
block0[0] = digest[0];
|
|
block0[1] = digest[1];
|
|
block0[2] = digest[2];
|
|
block0[3] = digest[3];
|
|
|
|
digest[0] = MD5M_A;
|
|
digest[1] = MD5M_B;
|
|
digest[2] = MD5M_C;
|
|
digest[3] = MD5M_D;
|
|
|
|
md5_transform (block0, block1, block2, block3, digest);
|
|
}
|
|
|
|
tmps[gid].digest_buf[0] = digest[0];
|
|
tmps[gid].digest_buf[1] = digest[1];
|
|
tmps[gid].digest_buf[2] = digest[2];
|
|
tmps[gid].digest_buf[3] = digest[3];
|
|
}
|
|
|
|
__kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m00400_comp (__global pw_t *pws, __global gpu_rule_t *rules_buf, __global comb_t *combs_buf, __global bf_t *bfs_buf, __global phpass_tmp_t *tmps, __global void *hooks, __global u32 *bitmaps_buf_s1_a, __global u32 *bitmaps_buf_s1_b, __global u32 *bitmaps_buf_s1_c, __global u32 *bitmaps_buf_s1_d, __global u32 *bitmaps_buf_s2_a, __global u32 *bitmaps_buf_s2_b, __global u32 *bitmaps_buf_s2_c, __global u32 *bitmaps_buf_s2_d, __global plain_t *plains_buf, __global digest_t *digests_buf, __global u32 *hashes_shown, __global salt_t *salt_bufs, __global void *esalt_bufs, __global u32 *d_return_buf, __global u32 *d_scryptV_buf, const u32 bitmap_mask, const u32 bitmap_shift1, const u32 bitmap_shift2, const u32 salt_pos, const u32 loop_pos, const u32 loop_cnt, const u32 rules_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
|
|
{
|
|
/**
|
|
* modifier
|
|
*/
|
|
|
|
const u32 gid = get_global_id (0);
|
|
|
|
if (gid >= gid_max) return;
|
|
|
|
const u32 lid = get_local_id (0);
|
|
|
|
/**
|
|
* digest
|
|
*/
|
|
|
|
const u32 r0 = tmps[gid].digest_buf[DGST_R0];
|
|
const u32 r1 = tmps[gid].digest_buf[DGST_R1];
|
|
const u32 r2 = tmps[gid].digest_buf[DGST_R2];
|
|
const u32 r3 = tmps[gid].digest_buf[DGST_R3];
|
|
|
|
#define il_pos 0
|
|
|
|
#include COMPARE_M
|
|
}
|