arno/hashcat
1
0
Fork 0
mirror of https://github.com/hashcat/hashcat.git synced 2024-11-26 09:58:16 +00:00
Code Issues Releases Wiki Activity

Cleanup -m 45xx kernels to latest standard

Browse Source
This commit is contained in:
Jens Steube 2016-04-10 17:40:51 +02:00
parent 0a180b6d64
commit 7948c74882
2 changed files with 182 additions and 232 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

40
OpenCL/m04500_a0.cl
View File

@ -39,7 +39,7 @@
__kernel void m04500_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, __global comb_t *combs_buf, __global bf_t *bfs_buf, __global void *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 il_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max) __kernel void m04500_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, __global comb_t *combs_buf, __global bf_t *bfs_buf, __global void *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 il_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
{ {
/** /**
* base * modifier
*/ */
const u32 gid = get_global_id (0); const u32 gid = get_global_id (0);
@ -70,18 +70,16 @@ __kernel void m04500_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf,
*/ */
u32 pw_buf0[4]; u32 pw_buf0[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];
u32 pw_buf1[4]; u32 pw_buf1[4];
pw_buf1[0] = pws[gid].i[ 4]; pw_buf0[0] = pws[gid].i[0];
pw_buf1[1] = pws[gid].i[ 5]; pw_buf0[1] = pws[gid].i[1];
pw_buf1[2] = pws[gid].i[ 6]; pw_buf0[2] = pws[gid].i[2];
pw_buf1[3] = pws[gid].i[ 7]; 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; const u32 pw_len = pws[gid].pw_len;
@ -378,7 +376,7 @@ __kernel void m04500_m16 (__global pw_t *pws, __global kernel_rule_t *rules_buf,
__kernel void m04500_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, __global comb_t *combs_buf, __global bf_t *bfs_buf, __global void *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 il_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max) __kernel void m04500_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, __global comb_t *combs_buf, __global bf_t *bfs_buf, __global void *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 il_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
{ {
/** /**
* base * modifier
*/ */
const u32 gid = get_global_id (0); const u32 gid = get_global_id (0);
@ -409,18 +407,16 @@ __kernel void m04500_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf,
*/ */
u32 pw_buf0[4]; u32 pw_buf0[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];
u32 pw_buf1[4]; u32 pw_buf1[4];
pw_buf1[0] = pws[gid].i[ 4]; pw_buf0[0] = pws[gid].i[0];
pw_buf1[1] = pws[gid].i[ 5]; pw_buf0[1] = pws[gid].i[1];
pw_buf1[2] = pws[gid].i[ 6]; pw_buf0[2] = pws[gid].i[2];
pw_buf1[3] = pws[gid].i[ 7]; 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; const u32 pw_len = pws[gid].pw_len;

374
OpenCL/m04500_a1.cl
View File

@ -5,6 +5,8 @@
#define _SHA1_ #define _SHA1_
#define NEW_SIMD_CODE
#include "include/constants.h" #include "include/constants.h"
#include "include/kernel_vendor.h" #include "include/kernel_vendor.h"
@ -16,16 +18,24 @@
#include "include/kernel_functions.c" #include "include/kernel_functions.c"
#include "OpenCL/types_ocl.c" #include "OpenCL/types_ocl.c"
#include "OpenCL/common.c" #include "OpenCL/common.c"
#include "OpenCL/simd.c"
#define COMPARE_S "OpenCL/check_single_comp4.c" #if VECT_SIZE == 1
#define COMPARE_M "OpenCL/check_multi_comp4.c" #define uint_to_hex_lower8_le(i) (u32x) (l_bin2asc[(i)])
#elif VECT_SIZE == 2
#define uint_to_hex_lower8_le(i) l_bin2asc[(i)] #define uint_to_hex_lower8_le(i) (u32x) (l_bin2asc[(i).s0], l_bin2asc[(i).s1])
#elif VECT_SIZE == 4
#define uint_to_hex_lower8_le(i) (u32x) (l_bin2asc[(i).s0], l_bin2asc[(i).s1], l_bin2asc[(i).s2], l_bin2asc[(i).s3])
#elif VECT_SIZE == 8
#define uint_to_hex_lower8_le(i) (u32x) (l_bin2asc[(i).s0], l_bin2asc[(i).s1], l_bin2asc[(i).s2], l_bin2asc[(i).s3], l_bin2asc[(i).s4], l_bin2asc[(i).s5], l_bin2asc[(i).s6], l_bin2asc[(i).s7])
#elif VECT_SIZE == 16
#define uint_to_hex_lower8_le(i) (u32x) (l_bin2asc[(i).s0], l_bin2asc[(i).s1], l_bin2asc[(i).s2], l_bin2asc[(i).s3], l_bin2asc[(i).s4], l_bin2asc[(i).s5], l_bin2asc[(i).s6], l_bin2asc[(i).s7], l_bin2asc[(i).s8], l_bin2asc[(i).s9], l_bin2asc[(i).sa], l_bin2asc[(i).sb], l_bin2asc[(i).sc], l_bin2asc[(i).sd], l_bin2asc[(i).se], l_bin2asc[(i).sf])
#endif
__kernel void m04500_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, __global comb_t *combs_buf, __global bf_t *bfs_buf, __global void *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 il_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max) __kernel void m04500_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, __global comb_t *combs_buf, __global bf_t *bfs_buf, __global void *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 il_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
{ {
/** /**
* base * modifier
*/ */
const u32 gid = get_global_id (0); const u32 gid = get_global_id (0);
@ -55,142 +65,119 @@ __kernel void m04500_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf,
* base * base
*/ */
u32 wordl0[4]; u32 pw_buf0[4];
u32 pw_buf1[4];
wordl0[0] = pws[gid].i[ 0]; pw_buf0[0] = pws[gid].i[0];
wordl0[1] = pws[gid].i[ 1]; pw_buf0[1] = pws[gid].i[1];
wordl0[2] = pws[gid].i[ 2]; pw_buf0[2] = pws[gid].i[2];
wordl0[3] = pws[gid].i[ 3]; pw_buf0[3] = pws[gid].i[3];
pw_buf1[0] = pws[gid].i[4];
u32 wordl1[4]; pw_buf1[1] = pws[gid].i[5];
pw_buf1[2] = pws[gid].i[6];
wordl1[0] = pws[gid].i[ 4]; pw_buf1[3] = pws[gid].i[7];
wordl1[1] = pws[gid].i[ 5];
wordl1[2] = pws[gid].i[ 6];
wordl1[3] = pws[gid].i[ 7];
u32 wordl2[4];
wordl2[0] = 0;
wordl2[1] = 0;
wordl2[2] = 0;
wordl2[3] = 0;
u32 wordl3[4];
wordl3[0] = 0;
wordl3[1] = 0;
wordl3[2] = 0;
wordl3[3] = 0;
const u32 pw_l_len = pws[gid].pw_len; const u32 pw_l_len = pws[gid].pw_len;
if (combs_mode == COMBINATOR_MODE_BASE_RIGHT)
{
append_0x80_2x4 (wordl0, wordl1, pw_l_len);
switch_buffer_by_offset_le (wordl0, wordl1, wordl2, wordl3, combs_buf[0].pw_len);
}
/** /**
* loop * loop
*/ */
for (u32 il_pos = 0; il_pos < il_cnt; il_pos++) for (u32 il_pos = 0; il_pos < il_cnt; il_pos += VECT_SIZE)
{ {
const u32 pw_r_len = combs_buf[il_pos].pw_len; const u32x pw_r_len = pwlenx_create_combt (combs_buf, il_pos);
const u32 pw_len = pw_l_len + pw_r_len; const u32x pw_len = pw_l_len + pw_r_len;
u32 wordr0[4]; /**
* concat password candidate
*/
wordr0[0] = combs_buf[il_pos].i[0]; u32x wordl0[4] = { 0 };
wordr0[1] = combs_buf[il_pos].i[1]; u32x wordl1[4] = { 0 };
wordr0[2] = combs_buf[il_pos].i[2]; u32x wordl2[4] = { 0 };
wordr0[3] = combs_buf[il_pos].i[3]; u32x wordl3[4] = { 0 };
u32 wordr1[4]; wordl0[0] = pw_buf0[0];
wordl0[1] = pw_buf0[1];
wordl0[2] = pw_buf0[2];
wordl0[3] = pw_buf0[3];
wordl1[0] = pw_buf1[0];
wordl1[1] = pw_buf1[1];
wordl1[2] = pw_buf1[2];
wordl1[3] = pw_buf1[3];
wordr1[0] = combs_buf[il_pos].i[4]; u32x wordr0[4] = { 0 };
wordr1[1] = combs_buf[il_pos].i[5]; u32x wordr1[4] = { 0 };
wordr1[2] = combs_buf[il_pos].i[6]; u32x wordr2[4] = { 0 };
wordr1[3] = combs_buf[il_pos].i[7]; u32x wordr3[4] = { 0 };
u32 wordr2[4]; wordr0[0] = ix_create_combt (combs_buf, il_pos, 0);
wordr0[1] = ix_create_combt (combs_buf, il_pos, 1);
wordr2[0] = 0; wordr0[2] = ix_create_combt (combs_buf, il_pos, 2);
wordr2[1] = 0; wordr0[3] = ix_create_combt (combs_buf, il_pos, 3);
wordr2[2] = 0; wordr1[0] = ix_create_combt (combs_buf, il_pos, 4);
wordr2[3] = 0; wordr1[1] = ix_create_combt (combs_buf, il_pos, 5);
wordr1[2] = ix_create_combt (combs_buf, il_pos, 6);
u32 wordr3[4]; wordr1[3] = ix_create_combt (combs_buf, il_pos, 7);
wordr3[0] = 0;
wordr3[1] = 0;
wordr3[2] = 0;
wordr3[3] = 0;
if (combs_mode == COMBINATOR_MODE_BASE_LEFT) if (combs_mode == COMBINATOR_MODE_BASE_LEFT)
{ {
append_0x80_2x4 (wordr0, wordr1, pw_r_len); switch_buffer_by_offset_le_VV (wordr0, wordr1, wordr2, wordr3, pw_l_len);
}
switch_buffer_by_offset_le (wordr0, wordr1, wordr2, wordr3, pw_l_len); else
{
switch_buffer_by_offset_le_VV (wordl0, wordl1, wordl2, wordl3, pw_r_len);
} }
u32 w0[4]; u32x w0[4];
u32x w1[4];
u32x w2[4];
u32x w3[4];
w0[0] = wordl0[0] | wordr0[0]; w0[0] = wordl0[0] | wordr0[0];
w0[1] = wordl0[1] | wordr0[1]; w0[1] = wordl0[1] | wordr0[1];
w0[2] = wordl0[2] | wordr0[2]; w0[2] = wordl0[2] | wordr0[2];
w0[3] = wordl0[3] | wordr0[3]; w0[3] = wordl0[3] | wordr0[3];
u32 w1[4];
w1[0] = wordl1[0] | wordr1[0]; w1[0] = wordl1[0] | wordr1[0];
w1[1] = wordl1[1] | wordr1[1]; w1[1] = wordl1[1] | wordr1[1];
w1[2] = wordl1[2] | wordr1[2]; w1[2] = wordl1[2] | wordr1[2];
w1[3] = wordl1[3] | wordr1[3]; w1[3] = wordl1[3] | wordr1[3];
u32 w2[4];
w2[0] = wordl2[0] | wordr2[0]; w2[0] = wordl2[0] | wordr2[0];
w2[1] = wordl2[1] | wordr2[1]; w2[1] = wordl2[1] | wordr2[1];
w2[2] = wordl2[2] | wordr2[2]; w2[2] = wordl2[2] | wordr2[2];
w2[3] = wordl2[3] | wordr2[3]; w2[3] = wordl2[3] | wordr2[3];
u32 w3[4];
w3[0] = wordl3[0] | wordr3[0]; w3[0] = wordl3[0] | wordr3[0];
w3[1] = wordl3[1] | wordr3[1]; w3[1] = wordl3[1] | wordr3[1];
w3[2] = 0; w3[2] = wordl3[2] | wordr3[2];
w3[3] = 0; w3[3] = wordl3[3] | wordr3[3];
/** /**
* sha1 * sha1
*/ */
u32 w0_t = swap32 (w0[0]); u32x w0_t = swap32 (w0[0]);
u32 w1_t = swap32 (w0[1]); u32x w1_t = swap32 (w0[1]);
u32 w2_t = swap32 (w0[2]); u32x w2_t = swap32 (w0[2]);
u32 w3_t = swap32 (w0[3]); u32x w3_t = swap32 (w0[3]);
u32 w4_t = swap32 (w1[0]); u32x w4_t = swap32 (w1[0]);
u32 w5_t = swap32 (w1[1]); u32x w5_t = swap32 (w1[1]);
u32 w6_t = swap32 (w1[2]); u32x w6_t = swap32 (w1[2]);
u32 w7_t = swap32 (w1[3]); u32x w7_t = swap32 (w1[3]);
u32 w8_t = swap32 (w2[0]); u32x w8_t = swap32 (w2[0]);
u32 w9_t = swap32 (w2[1]); u32x w9_t = swap32 (w2[1]);
u32 wa_t = swap32 (w2[2]); u32x wa_t = swap32 (w2[2]);
u32 wb_t = swap32 (w2[3]); u32x wb_t = swap32 (w2[3]);
u32 wc_t = swap32 (w3[0]); u32x wc_t = swap32 (w3[0]);
u32 wd_t = swap32 (w3[1]); u32x wd_t = swap32 (w3[1]);
u32 we_t = 0; u32x we_t = 0;
u32 wf_t = pw_len * 8; u32x wf_t = pw_len * 8;
u32 a = SHA1M_A; u32x a = SHA1M_A;
u32 b = SHA1M_B; u32x b = SHA1M_B;
u32 c = SHA1M_C; u32x c = SHA1M_C;
u32 d = SHA1M_D; u32x d = SHA1M_D;
u32 e = SHA1M_E; u32x e = SHA1M_E;
#undef K #undef K
#define K SHA1C00 #define K SHA1C00
@ -428,12 +415,7 @@ __kernel void m04500_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf,
we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, c, d, e, a, b, we_t); we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, c, d, e, a, b, we_t);
wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, b, c, d, e, a, wf_t); wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, b, c, d, e, a, wf_t);
const u32 r0 = d; COMPARE_M_SIMD (d, e, c, b);
const u32 r1 = e;
const u32 r2 = c;
const u32 r3 = b;
#include COMPARE_M
} }
} }
@ -448,7 +430,7 @@ __kernel void m04500_m16 (__global pw_t *pws, __global kernel_rule_t *rules_buf,
__kernel void m04500_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, __global comb_t *combs_buf, __global bf_t *bfs_buf, __global void *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 il_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max) __kernel void m04500_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, __global comb_t *combs_buf, __global bf_t *bfs_buf, __global void *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 il_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
{ {
/** /**
* base * modifier
*/ */
const u32 gid = get_global_id (0); const u32 gid = get_global_id (0);
@ -478,43 +460,20 @@ __kernel void m04500_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf,
* base * base
*/ */
u32 wordl0[4]; u32 pw_buf0[4];
u32 pw_buf1[4];
wordl0[0] = pws[gid].i[ 0]; pw_buf0[0] = pws[gid].i[0];
wordl0[1] = pws[gid].i[ 1]; pw_buf0[1] = pws[gid].i[1];
wordl0[2] = pws[gid].i[ 2]; pw_buf0[2] = pws[gid].i[2];
wordl0[3] = pws[gid].i[ 3]; pw_buf0[3] = pws[gid].i[3];
pw_buf1[0] = pws[gid].i[4];
u32 wordl1[4]; pw_buf1[1] = pws[gid].i[5];
pw_buf1[2] = pws[gid].i[6];
wordl1[0] = pws[gid].i[ 4]; pw_buf1[3] = pws[gid].i[7];
wordl1[1] = pws[gid].i[ 5];
wordl1[2] = pws[gid].i[ 6];
wordl1[3] = pws[gid].i[ 7];
u32 wordl2[4];
wordl2[0] = 0;
wordl2[1] = 0;
wordl2[2] = 0;
wordl2[3] = 0;
u32 wordl3[4];
wordl3[0] = 0;
wordl3[1] = 0;
wordl3[2] = 0;
wordl3[3] = 0;
const u32 pw_l_len = pws[gid].pw_len; const u32 pw_l_len = pws[gid].pw_len;
if (combs_mode == COMBINATOR_MODE_BASE_RIGHT)
{
append_0x80_2x4 (wordl0, wordl1, pw_l_len);
switch_buffer_by_offset_le (wordl0, wordl1, wordl2, wordl3, combs_buf[0].pw_len);
}
/** /**
* digest * digest
*/ */
@ -531,107 +490,107 @@ __kernel void m04500_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf,
* reverse * reverse
*/ */
const u32 e_rev = rotl32 (search[1], 2u); const u32 e_rev = rotl32_S (search[1], 2u);
/** /**
* loop * loop
*/ */
for (u32 il_pos = 0; il_pos < il_cnt; il_pos++) for (u32 il_pos = 0; il_pos < il_cnt; il_pos += VECT_SIZE)
{ {
const u32 pw_r_len = combs_buf[il_pos].pw_len; const u32x pw_r_len = pwlenx_create_combt (combs_buf, il_pos);
const u32 pw_len = pw_l_len + pw_r_len; const u32x pw_len = pw_l_len + pw_r_len;
u32 wordr0[4]; /**
* concat password candidate
*/
wordr0[0] = combs_buf[il_pos].i[0]; u32x wordl0[4] = { 0 };
wordr0[1] = combs_buf[il_pos].i[1]; u32x wordl1[4] = { 0 };
wordr0[2] = combs_buf[il_pos].i[2]; u32x wordl2[4] = { 0 };
wordr0[3] = combs_buf[il_pos].i[3]; u32x wordl3[4] = { 0 };
u32 wordr1[4]; wordl0[0] = pw_buf0[0];
wordl0[1] = pw_buf0[1];
wordl0[2] = pw_buf0[2];
wordl0[3] = pw_buf0[3];
wordl1[0] = pw_buf1[0];
wordl1[1] = pw_buf1[1];
wordl1[2] = pw_buf1[2];
wordl1[3] = pw_buf1[3];
wordr1[0] = combs_buf[il_pos].i[4]; u32x wordr0[4] = { 0 };
wordr1[1] = combs_buf[il_pos].i[5]; u32x wordr1[4] = { 0 };
wordr1[2] = combs_buf[il_pos].i[6]; u32x wordr2[4] = { 0 };
wordr1[3] = combs_buf[il_pos].i[7]; u32x wordr3[4] = { 0 };
u32 wordr2[4]; wordr0[0] = ix_create_combt (combs_buf, il_pos, 0);
wordr0[1] = ix_create_combt (combs_buf, il_pos, 1);
wordr2[0] = 0; wordr0[2] = ix_create_combt (combs_buf, il_pos, 2);
wordr2[1] = 0; wordr0[3] = ix_create_combt (combs_buf, il_pos, 3);
wordr2[2] = 0; wordr1[0] = ix_create_combt (combs_buf, il_pos, 4);
wordr2[3] = 0; wordr1[1] = ix_create_combt (combs_buf, il_pos, 5);
wordr1[2] = ix_create_combt (combs_buf, il_pos, 6);
u32 wordr3[4]; wordr1[3] = ix_create_combt (combs_buf, il_pos, 7);
wordr3[0] = 0;
wordr3[1] = 0;
wordr3[2] = 0;
wordr3[3] = 0;
if (combs_mode == COMBINATOR_MODE_BASE_LEFT) if (combs_mode == COMBINATOR_MODE_BASE_LEFT)
{ {
append_0x80_2x4 (wordr0, wordr1, pw_r_len); switch_buffer_by_offset_le_VV (wordr0, wordr1, wordr2, wordr3, pw_l_len);
}
switch_buffer_by_offset_le (wordr0, wordr1, wordr2, wordr3, pw_l_len); else
{
switch_buffer_by_offset_le_VV (wordl0, wordl1, wordl2, wordl3, pw_r_len);
} }
u32 w0[4]; u32x w0[4];
u32x w1[4];
u32x w2[4];
u32x w3[4];
w0[0] = wordl0[0] | wordr0[0]; w0[0] = wordl0[0] | wordr0[0];
w0[1] = wordl0[1] | wordr0[1]; w0[1] = wordl0[1] | wordr0[1];
w0[2] = wordl0[2] | wordr0[2]; w0[2] = wordl0[2] | wordr0[2];
w0[3] = wordl0[3] | wordr0[3]; w0[3] = wordl0[3] | wordr0[3];
u32 w1[4];
w1[0] = wordl1[0] | wordr1[0]; w1[0] = wordl1[0] | wordr1[0];
w1[1] = wordl1[1] | wordr1[1]; w1[1] = wordl1[1] | wordr1[1];
w1[2] = wordl1[2] | wordr1[2]; w1[2] = wordl1[2] | wordr1[2];
w1[3] = wordl1[3] | wordr1[3]; w1[3] = wordl1[3] | wordr1[3];
u32 w2[4];
w2[0] = wordl2[0] | wordr2[0]; w2[0] = wordl2[0] | wordr2[0];
w2[1] = wordl2[1] | wordr2[1]; w2[1] = wordl2[1] | wordr2[1];
w2[2] = wordl2[2] | wordr2[2]; w2[2] = wordl2[2] | wordr2[2];
w2[3] = wordl2[3] | wordr2[3]; w2[3] = wordl2[3] | wordr2[3];
u32 w3[4];
w3[0] = wordl3[0] | wordr3[0]; w3[0] = wordl3[0] | wordr3[0];
w3[1] = wordl3[1] | wordr3[1]; w3[1] = wordl3[1] | wordr3[1];
w3[2] = 0; w3[2] = wordl3[2] | wordr3[2];
w3[3] = 0; w3[3] = wordl3[3] | wordr3[3];
/** /**
* sha1 * sha1
*/ */
u32 w0_t = swap32 (w0[0]); u32x w0_t = swap32 (w0[0]);
u32 w1_t = swap32 (w0[1]); u32x w1_t = swap32 (w0[1]);
u32 w2_t = swap32 (w0[2]); u32x w2_t = swap32 (w0[2]);
u32 w3_t = swap32 (w0[3]); u32x w3_t = swap32 (w0[3]);
u32 w4_t = swap32 (w1[0]); u32x w4_t = swap32 (w1[0]);
u32 w5_t = swap32 (w1[1]); u32x w5_t = swap32 (w1[1]);
u32 w6_t = swap32 (w1[2]); u32x w6_t = swap32 (w1[2]);
u32 w7_t = swap32 (w1[3]); u32x w7_t = swap32 (w1[3]);
u32 w8_t = swap32 (w2[0]); u32x w8_t = swap32 (w2[0]);
u32 w9_t = swap32 (w2[1]); u32x w9_t = swap32 (w2[1]);
u32 wa_t = swap32 (w2[2]); u32x wa_t = swap32 (w2[2]);
u32 wb_t = swap32 (w2[3]); u32x wb_t = swap32 (w2[3]);
u32 wc_t = swap32 (w3[0]); u32x wc_t = swap32 (w3[0]);
u32 wd_t = swap32 (w3[1]); u32x wd_t = swap32 (w3[1]);
u32 we_t = 0; u32x we_t = 0;
u32 wf_t = pw_len * 8; u32x wf_t = pw_len * 8;
u32 a = SHA1M_A; u32x a = SHA1M_A;
u32 b = SHA1M_B; u32x b = SHA1M_B;
u32 c = SHA1M_C; u32x c = SHA1M_C;
u32 d = SHA1M_D; u32x d = SHA1M_D;
u32 e = SHA1M_E; u32x e = SHA1M_E;
#undef K #undef K
#define K SHA1C00 #define K SHA1C00
@ -865,19 +824,14 @@ __kernel void m04500_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf,
wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F1, b, c, d, e, a, wa_t); wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F1, b, c, d, e, a, wa_t);
wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F1, a, b, c, d, e, wb_t); wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F1, a, b, c, d, e, wb_t);
if (allx (e != e_rev)) continue; if (MATCHES_NONE_VS (e, e_rev)) continue;
wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F1, e, a, b, c, d, wc_t); wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F1, e, a, b, c, d, wc_t);
wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F1, d, e, a, b, c, wd_t); wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F1, d, e, a, b, c, wd_t);
we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, c, d, e, a, b, we_t); we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, c, d, e, a, b, we_t);
wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, b, c, d, e, a, wf_t); wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, b, c, d, e, a, wf_t);
const u32 r0 = d; COMPARE_S_SIMD (d, e, c, b);
const u32 r1 = e;
const u32 r2 = c;
const u32 r3 = b;
#include COMPARE_S
} }
} }