arno/hashcat
1
0
Fork 0
mirror of https://github.com/hashcat/hashcat.git synced 2025-07-29 09:59:06 +00:00
Code Issues Releases Wiki Activity

Update module for the Assimilation Bridge & Convert newlines

Browse Source
This commit is contained in:
PenguinKeeper7 2025-05-31 00:53:09 +01:00
parent 4cb22387b3
commit d7ef1aa3a0
7 changed files with 4007 additions and 4005 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1740
OpenCL/m32800_a0-optimized.cl
View File

File diff suppressed because it is too large Load Diff

614
OpenCL/m32800_a0-pure.cl
View File

@ -1,307 +1,307 @@
/** /**
* Author......: See docs/credits.txt * Author......: See docs/credits.txt
* License.....: MIT * License.....: MIT
*/ */
//#define NEW_SIMD_CODE //#define NEW_SIMD_CODE
#ifdef KERNEL_STATIC #ifdef KERNEL_STATIC
#include M2S(INCLUDE_PATH/inc_vendor.h) #include M2S(INCLUDE_PATH/inc_vendor.h)
#include M2S(INCLUDE_PATH/inc_types.h) #include M2S(INCLUDE_PATH/inc_types.h)
#include M2S(INCLUDE_PATH/inc_platform.cl) #include M2S(INCLUDE_PATH/inc_platform.cl)
#include M2S(INCLUDE_PATH/inc_common.cl) #include M2S(INCLUDE_PATH/inc_common.cl)
#include M2S(INCLUDE_PATH/inc_rp.h) #include M2S(INCLUDE_PATH/inc_rp.h)
#include M2S(INCLUDE_PATH/inc_rp.cl) #include M2S(INCLUDE_PATH/inc_rp.cl)
#include M2S(INCLUDE_PATH/inc_scalar.cl) #include M2S(INCLUDE_PATH/inc_scalar.cl)
#include M2S(INCLUDE_PATH/inc_hash_md5.cl) #include M2S(INCLUDE_PATH/inc_hash_md5.cl)
#include M2S(INCLUDE_PATH/inc_hash_sha1.cl) #include M2S(INCLUDE_PATH/inc_hash_sha1.cl)
#endif #endif
#if VECT_SIZE == 1 #if VECT_SIZE == 1
#define uint_to_hex_lower8(i) make_u32x (b_bin2asc[(i)]) #define uint_to_hex_lower8(i) make_u32x (b_bin2asc[(i)])
#define uint_to_hex_lower8_le(i) make_u32x (l_bin2asc[(i)]) #define uint_to_hex_lower8_le(i) make_u32x (l_bin2asc[(i)])
#elif VECT_SIZE == 2 #elif VECT_SIZE == 2
#define uint_to_hex_lower8(i) make_u32x (b_bin2asc[(i).s0], b_bin2asc[(i).s1]) #define uint_to_hex_lower8(i) make_u32x (b_bin2asc[(i).s0], b_bin2asc[(i).s1])
#define uint_to_hex_lower8_le(i) make_u32x (l_bin2asc[(i).s0], l_bin2asc[(i).s1]) #define uint_to_hex_lower8_le(i) make_u32x (l_bin2asc[(i).s0], l_bin2asc[(i).s1])
#elif VECT_SIZE == 4 #elif VECT_SIZE == 4
#define uint_to_hex_lower8(i) make_u32x (b_bin2asc[(i).s0], b_bin2asc[(i).s1], b_bin2asc[(i).s2], b_bin2asc[(i).s3]) #define uint_to_hex_lower8(i) make_u32x (b_bin2asc[(i).s0], b_bin2asc[(i).s1], b_bin2asc[(i).s2], b_bin2asc[(i).s3])
#define uint_to_hex_lower8_le(i) make_u32x (l_bin2asc[(i).s0], l_bin2asc[(i).s1], l_bin2asc[(i).s2], l_bin2asc[(i).s3]) #define uint_to_hex_lower8_le(i) make_u32x (l_bin2asc[(i).s0], l_bin2asc[(i).s1], l_bin2asc[(i).s2], l_bin2asc[(i).s3])
#elif VECT_SIZE == 8 #elif VECT_SIZE == 8
#define uint_to_hex_lower8(i) make_u32x (b_bin2asc[(i).s0], b_bin2asc[(i).s1], b_bin2asc[(i).s2], b_bin2asc[(i).s3], b_bin2asc[(i).s4], b_bin2asc[(i).s5], b_bin2asc[(i).s6], b_bin2asc[(i).s7]) #define uint_to_hex_lower8(i) make_u32x (b_bin2asc[(i).s0], b_bin2asc[(i).s1], b_bin2asc[(i).s2], b_bin2asc[(i).s3], b_bin2asc[(i).s4], b_bin2asc[(i).s5], b_bin2asc[(i).s6], b_bin2asc[(i).s7])
#define uint_to_hex_lower8_le(i) make_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]) #define uint_to_hex_lower8_le(i) make_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 #elif VECT_SIZE == 16
#define uint_to_hex_lower8(i) make_u32x (b_bin2asc[(i).s0], b_bin2asc[(i).s1], b_bin2asc[(i).s2], b_bin2asc[(i).s3], b_bin2asc[(i).s4], b_bin2asc[(i).s5], b_bin2asc[(i).s6], b_bin2asc[(i).s7], b_bin2asc[(i).s8], b_bin2asc[(i).s9], b_bin2asc[(i).sa], b_bin2asc[(i).sb], b_bin2asc[(i).sc], b_bin2asc[(i).sd], b_bin2asc[(i).se], b_bin2asc[(i).sf]) #define uint_to_hex_lower8(i) make_u32x (b_bin2asc[(i).s0], b_bin2asc[(i).s1], b_bin2asc[(i).s2], b_bin2asc[(i).s3], b_bin2asc[(i).s4], b_bin2asc[(i).s5], b_bin2asc[(i).s6], b_bin2asc[(i).s7], b_bin2asc[(i).s8], b_bin2asc[(i).s9], b_bin2asc[(i).sa], b_bin2asc[(i).sb], b_bin2asc[(i).sc], b_bin2asc[(i).sd], b_bin2asc[(i).se], b_bin2asc[(i).sf])
#define uint_to_hex_lower8_le(i) make_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]) #define uint_to_hex_lower8_le(i) make_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 #endif
KERNEL_FQ void m32800_mxx (KERN_ATTR_RULES ()) KERNEL_FQ void m32800_mxx (KERN_ATTR_RULES ())
{ {
/** /**
* modifier * modifier
*/ */
const u64 gid = get_global_id (0); const u64 gid = get_global_id (0);
const u64 lid = get_local_id (0); const u64 lid = get_local_id (0);
const u64 lsz = get_local_size (0); const u64 lsz = get_local_size (0);
/** /**
* bin2asc tables * bin2asc tables
*/ */
LOCAL_VK u32 l_bin2asc[256]; LOCAL_VK u32 l_bin2asc[256];
LOCAL_VK u32 b_bin2asc[256]; LOCAL_VK u32 b_bin2asc[256];
for (u32 i = lid; i < 256; i += lsz) for (u32 i = lid; i < 256; i += lsz)
{ {
const u32 i0 = (i >> 0) & 15; const u32 i0 = (i >> 0) & 15;
const u32 i1 = (i >> 4) & 15; const u32 i1 = (i >> 4) & 15;
l_bin2asc[i] = ((i0 < 10) ? '0' + i0 : 'a' - 10 + i0) << 0 l_bin2asc[i] = ((i0 < 10) ? '0' + i0 : 'a' - 10 + i0) << 0
| ((i1 < 10) ? '0' + i1 : 'a' - 10 + i1) << 8; | ((i1 < 10) ? '0' + i1 : 'a' - 10 + i1) << 8;
b_bin2asc[i] = ((i0 < 10) ? '0' + i0 : 'a' - 10 + i0) << 8 b_bin2asc[i] = ((i0 < 10) ? '0' + i0 : 'a' - 10 + i0) << 8
| ((i1 < 10) ? '0' + i1 : 'a' - 10 + i1) << 0; | ((i1 < 10) ? '0' + i1 : 'a' - 10 + i1) << 0;
} }
SYNC_THREADS (); SYNC_THREADS ();
if (gid >= GID_CNT) return; if (gid >= GID_CNT) return;
/** /**
* base * base
*/ */
COPY_PW (pws[gid]); COPY_PW (pws[gid]);
/** /**
* loop * loop
*/ */
for (u32 il_pos = 0; il_pos < IL_CNT; il_pos++) for (u32 il_pos = 0; il_pos < IL_CNT; il_pos++)
{ {
pw_t tmp = PASTE_PW; pw_t tmp = PASTE_PW;
tmp.pw_len = apply_rules (rules_buf[il_pos].cmds, tmp.i, tmp.pw_len); tmp.pw_len = apply_rules (rules_buf[il_pos].cmds, tmp.i, tmp.pw_len);
md5_ctx_t ctx0; md5_ctx_t ctx0;
md5_init (&ctx0); md5_init (&ctx0);
md5_update (&ctx0, tmp.i, tmp.pw_len); md5_update (&ctx0, tmp.i, tmp.pw_len);
md5_final (&ctx0); md5_final (&ctx0);
u32 a = ctx0.h[0]; u32 a = ctx0.h[0];
u32 b = ctx0.h[1]; u32 b = ctx0.h[1];
u32 c = ctx0.h[2]; u32 c = ctx0.h[2];
u32 d = ctx0.h[3]; u32 d = ctx0.h[3];
sha1_ctx_t ctx1; sha1_ctx_t ctx1;
sha1_init (&ctx1); sha1_init (&ctx1);
ctx1.w0[0] = uint_to_hex_lower8_le ((a >> 8) & 255) << 0 ctx1.w0[0] = uint_to_hex_lower8_le ((a >> 8) & 255) << 0
| uint_to_hex_lower8_le ((a >> 0) & 255) << 16; | uint_to_hex_lower8_le ((a >> 0) & 255) << 16;
ctx1.w0[1] = uint_to_hex_lower8_le ((a >> 24) & 255) << 0 ctx1.w0[1] = uint_to_hex_lower8_le ((a >> 24) & 255) << 0
| uint_to_hex_lower8_le ((a >> 16) & 255) << 16; | uint_to_hex_lower8_le ((a >> 16) & 255) << 16;
ctx1.w0[2] = uint_to_hex_lower8_le ((b >> 8) & 255) << 0 ctx1.w0[2] = uint_to_hex_lower8_le ((b >> 8) & 255) << 0
| uint_to_hex_lower8_le ((b >> 0) & 255) << 16; | uint_to_hex_lower8_le ((b >> 0) & 255) << 16;
ctx1.w0[3] = uint_to_hex_lower8_le ((b >> 24) & 255) << 0 ctx1.w0[3] = uint_to_hex_lower8_le ((b >> 24) & 255) << 0
| uint_to_hex_lower8_le ((b >> 16) & 255) << 16; | uint_to_hex_lower8_le ((b >> 16) & 255) << 16;
ctx1.w1[0] = uint_to_hex_lower8_le ((c >> 8) & 255) << 0 ctx1.w1[0] = uint_to_hex_lower8_le ((c >> 8) & 255) << 0
| uint_to_hex_lower8_le ((c >> 0) & 255) << 16; | uint_to_hex_lower8_le ((c >> 0) & 255) << 16;
ctx1.w1[1] = uint_to_hex_lower8_le ((c >> 24) & 255) << 0 ctx1.w1[1] = uint_to_hex_lower8_le ((c >> 24) & 255) << 0
| uint_to_hex_lower8_le ((c >> 16) & 255) << 16; | uint_to_hex_lower8_le ((c >> 16) & 255) << 16;
ctx1.w1[2] = uint_to_hex_lower8_le ((d >> 8) & 255) << 0 ctx1.w1[2] = uint_to_hex_lower8_le ((d >> 8) & 255) << 0
| uint_to_hex_lower8_le ((d >> 0) & 255) << 16; | uint_to_hex_lower8_le ((d >> 0) & 255) << 16;
ctx1.w1[3] = uint_to_hex_lower8_le ((d >> 24) & 255) << 0 ctx1.w1[3] = uint_to_hex_lower8_le ((d >> 24) & 255) << 0
| uint_to_hex_lower8_le ((d >> 16) & 255) << 16; | uint_to_hex_lower8_le ((d >> 16) & 255) << 16;
ctx1.len = 32; ctx1.len = 32;
sha1_final (&ctx1); sha1_final (&ctx1);
a = ctx1.h[0]; a = ctx1.h[0];
b = ctx1.h[1]; b = ctx1.h[1];
c = ctx1.h[2]; c = ctx1.h[2];
d = ctx1.h[3]; d = ctx1.h[3];
const u32 e = ctx1.h[4]; const u32 e = ctx1.h[4];
md5_ctx_t ctx2; md5_ctx_t ctx2;
md5_init (&ctx2); md5_init (&ctx2);
ctx2.w0[0] = uint_to_hex_lower8 ((a >> 24) & 255) << 0 ctx2.w0[0] = uint_to_hex_lower8 ((a >> 24) & 255) << 0
| uint_to_hex_lower8 ((a >> 16) & 255) << 16; | uint_to_hex_lower8 ((a >> 16) & 255) << 16;
ctx2.w0[1] = uint_to_hex_lower8 ((a >> 8) & 255) << 0 ctx2.w0[1] = uint_to_hex_lower8 ((a >> 8) & 255) << 0
| uint_to_hex_lower8 ((a >> 0) & 255) << 16; | uint_to_hex_lower8 ((a >> 0) & 255) << 16;
ctx2.w0[2] = uint_to_hex_lower8 ((b >> 24) & 255) << 0 ctx2.w0[2] = uint_to_hex_lower8 ((b >> 24) & 255) << 0
| uint_to_hex_lower8 ((b >> 16) & 255) << 16; | uint_to_hex_lower8 ((b >> 16) & 255) << 16;
ctx2.w0[3] = uint_to_hex_lower8 ((b >> 8) & 255) << 0 ctx2.w0[3] = uint_to_hex_lower8 ((b >> 8) & 255) << 0
| uint_to_hex_lower8 ((b >> 0) & 255) << 16; | uint_to_hex_lower8 ((b >> 0) & 255) << 16;
ctx2.w1[0] = uint_to_hex_lower8 ((c >> 24) & 255) << 0 ctx2.w1[0] = uint_to_hex_lower8 ((c >> 24) & 255) << 0
| uint_to_hex_lower8 ((c >> 16) & 255) << 16; | uint_to_hex_lower8 ((c >> 16) & 255) << 16;
ctx2.w1[1] = uint_to_hex_lower8 ((c >> 8) & 255) << 0 ctx2.w1[1] = uint_to_hex_lower8 ((c >> 8) & 255) << 0
| uint_to_hex_lower8 ((c >> 0) & 255) << 16; | uint_to_hex_lower8 ((c >> 0) & 255) << 16;
ctx2.w1[2] = uint_to_hex_lower8 ((d >> 24) & 255) << 0 ctx2.w1[2] = uint_to_hex_lower8 ((d >> 24) & 255) << 0
| uint_to_hex_lower8 ((d >> 16) & 255) << 16; | uint_to_hex_lower8 ((d >> 16) & 255) << 16;
ctx2.w1[3] = uint_to_hex_lower8 ((d >> 8) & 255) << 0 ctx2.w1[3] = uint_to_hex_lower8 ((d >> 8) & 255) << 0
| uint_to_hex_lower8 ((d >> 0) & 255) << 16; | uint_to_hex_lower8 ((d >> 0) & 255) << 16;
ctx2.w2[0] = uint_to_hex_lower8 ((e >> 24) & 255) << 0 ctx2.w2[0] = uint_to_hex_lower8 ((e >> 24) & 255) << 0
| uint_to_hex_lower8 ((e >> 16) & 255) << 16; | uint_to_hex_lower8 ((e >> 16) & 255) << 16;
ctx2.w2[1] = uint_to_hex_lower8 ((e >> 8) & 255) << 0 ctx2.w2[1] = uint_to_hex_lower8 ((e >> 8) & 255) << 0
| uint_to_hex_lower8 ((e >> 0) & 255) << 16; | uint_to_hex_lower8 ((e >> 0) & 255) << 16;
ctx2.len = 40; ctx2.len = 40;
md5_final (&ctx2); md5_final (&ctx2);
const u32 r0 = ctx2.h[DGST_R0]; const u32 r0 = ctx2.h[DGST_R0];
const u32 r1 = ctx2.h[DGST_R1]; const u32 r1 = ctx2.h[DGST_R1];
const u32 r2 = ctx2.h[DGST_R2]; const u32 r2 = ctx2.h[DGST_R2];
const u32 r3 = ctx2.h[DGST_R3]; const u32 r3 = ctx2.h[DGST_R3];
COMPARE_M_SCALAR (r0, r1, r2, r3); COMPARE_M_SCALAR (r0, r1, r2, r3);
} }
} }
KERNEL_FQ void m32800_sxx (KERN_ATTR_RULES ()) KERNEL_FQ void m32800_sxx (KERN_ATTR_RULES ())
{ {
/** /**
* modifier * modifier
*/ */
const u64 gid = get_global_id (0); const u64 gid = get_global_id (0);
const u64 lid = get_local_id (0); const u64 lid = get_local_id (0);
const u64 lsz = get_local_size (0); const u64 lsz = get_local_size (0);
/** /**
* bin2asc tables * bin2asc tables
*/ */
LOCAL_VK u32 l_bin2asc[256]; LOCAL_VK u32 l_bin2asc[256];
LOCAL_VK u32 b_bin2asc[256]; LOCAL_VK u32 b_bin2asc[256];
for (u32 i = lid; i < 256; i += lsz) for (u32 i = lid; i < 256; i += lsz)
{ {
const u32 i0 = (i >> 0) & 15; const u32 i0 = (i >> 0) & 15;
const u32 i1 = (i >> 4) & 15; const u32 i1 = (i >> 4) & 15;
l_bin2asc[i] = ((i0 < 10) ? '0' + i0 : 'a' - 10 + i0) << 0 l_bin2asc[i] = ((i0 < 10) ? '0' + i0 : 'a' - 10 + i0) << 0
| ((i1 < 10) ? '0' + i1 : 'a' - 10 + i1) << 8; | ((i1 < 10) ? '0' + i1 : 'a' - 10 + i1) << 8;
b_bin2asc[i] = ((i0 < 10) ? '0' + i0 : 'a' - 10 + i0) << 8 b_bin2asc[i] = ((i0 < 10) ? '0' + i0 : 'a' - 10 + i0) << 8
| ((i1 < 10) ? '0' + i1 : 'a' - 10 + i1) << 0; | ((i1 < 10) ? '0' + i1 : 'a' - 10 + i1) << 0;
} }
SYNC_THREADS (); SYNC_THREADS ();
if (gid >= GID_CNT) return; if (gid >= GID_CNT) return;
/** /**
* digest * digest
*/ */
const u32 search[4] = const u32 search[4] =
{ {
digests_buf[DIGESTS_OFFSET_HOST].digest_buf[DGST_R0], 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_R1],
digests_buf[DIGESTS_OFFSET_HOST].digest_buf[DGST_R2], digests_buf[DIGESTS_OFFSET_HOST].digest_buf[DGST_R2],
digests_buf[DIGESTS_OFFSET_HOST].digest_buf[DGST_R3] digests_buf[DIGESTS_OFFSET_HOST].digest_buf[DGST_R3]
}; };
/** /**
* base * base
*/ */
COPY_PW (pws[gid]); COPY_PW (pws[gid]);
/** /**
* loop * loop
*/ */
for (u32 il_pos = 0; il_pos < IL_CNT; il_pos++) for (u32 il_pos = 0; il_pos < IL_CNT; il_pos++)
{ {
pw_t tmp = PASTE_PW; pw_t tmp = PASTE_PW;
tmp.pw_len = apply_rules (rules_buf[il_pos].cmds, tmp.i, tmp.pw_len); tmp.pw_len = apply_rules (rules_buf[il_pos].cmds, tmp.i, tmp.pw_len);
md5_ctx_t ctx0; md5_ctx_t ctx0;
md5_init (&ctx0); md5_init (&ctx0);
md5_update (&ctx0, tmp.i, tmp.pw_len); md5_update (&ctx0, tmp.i, tmp.pw_len);
md5_final (&ctx0); md5_final (&ctx0);
u32 a = ctx0.h[0]; u32 a = ctx0.h[0];
u32 b = ctx0.h[1]; u32 b = ctx0.h[1];
u32 c = ctx0.h[2]; u32 c = ctx0.h[2];
u32 d = ctx0.h[3]; u32 d = ctx0.h[3];
sha1_ctx_t ctx1; sha1_ctx_t ctx1;
sha1_init (&ctx1); sha1_init (&ctx1);
ctx1.w0[0] = uint_to_hex_lower8_le ((a >> 8) & 255) << 0 ctx1.w0[0] = uint_to_hex_lower8_le ((a >> 8) & 255) << 0
| uint_to_hex_lower8_le ((a >> 0) & 255) << 16; | uint_to_hex_lower8_le ((a >> 0) & 255) << 16;
ctx1.w0[1] = uint_to_hex_lower8_le ((a >> 24) & 255) << 0 ctx1.w0[1] = uint_to_hex_lower8_le ((a >> 24) & 255) << 0
| uint_to_hex_lower8_le ((a >> 16) & 255) << 16; | uint_to_hex_lower8_le ((a >> 16) & 255) << 16;
ctx1.w0[2] = uint_to_hex_lower8_le ((b >> 8) & 255) << 0 ctx1.w0[2] = uint_to_hex_lower8_le ((b >> 8) & 255) << 0
| uint_to_hex_lower8_le ((b >> 0) & 255) << 16; | uint_to_hex_lower8_le ((b >> 0) & 255) << 16;
ctx1.w0[3] = uint_to_hex_lower8_le ((b >> 24) & 255) << 0 ctx1.w0[3] = uint_to_hex_lower8_le ((b >> 24) & 255) << 0
| uint_to_hex_lower8_le ((b >> 16) & 255) << 16; | uint_to_hex_lower8_le ((b >> 16) & 255) << 16;
ctx1.w1[0] = uint_to_hex_lower8_le ((c >> 8) & 255) << 0 ctx1.w1[0] = uint_to_hex_lower8_le ((c >> 8) & 255) << 0
| uint_to_hex_lower8_le ((c >> 0) & 255) << 16; | uint_to_hex_lower8_le ((c >> 0) & 255) << 16;
ctx1.w1[1] = uint_to_hex_lower8_le ((c >> 24) & 255) << 0 ctx1.w1[1] = uint_to_hex_lower8_le ((c >> 24) & 255) << 0
| uint_to_hex_lower8_le ((c >> 16) & 255) << 16; | uint_to_hex_lower8_le ((c >> 16) & 255) << 16;
ctx1.w1[2] = uint_to_hex_lower8_le ((d >> 8) & 255) << 0 ctx1.w1[2] = uint_to_hex_lower8_le ((d >> 8) & 255) << 0
| uint_to_hex_lower8_le ((d >> 0) & 255) << 16; | uint_to_hex_lower8_le ((d >> 0) & 255) << 16;
ctx1.w1[3] = uint_to_hex_lower8_le ((d >> 24) & 255) << 0 ctx1.w1[3] = uint_to_hex_lower8_le ((d >> 24) & 255) << 0
| uint_to_hex_lower8_le ((d >> 16) & 255) << 16; | uint_to_hex_lower8_le ((d >> 16) & 255) << 16;
ctx1.len = 32; ctx1.len = 32;
sha1_final (&ctx1); sha1_final (&ctx1);
a = ctx1.h[0]; a = ctx1.h[0];
b = ctx1.h[1]; b = ctx1.h[1];
c = ctx1.h[2]; c = ctx1.h[2];
d = ctx1.h[3]; d = ctx1.h[3];
const u32 e = ctx1.h[4]; const u32 e = ctx1.h[4];
md5_ctx_t ctx2; md5_ctx_t ctx2;
md5_init (&ctx2); md5_init (&ctx2);
ctx2.w0[0] = uint_to_hex_lower8 ((a >> 24) & 255) << 0 ctx2.w0[0] = uint_to_hex_lower8 ((a >> 24) & 255) << 0
| uint_to_hex_lower8 ((a >> 16) & 255) << 16; | uint_to_hex_lower8 ((a >> 16) & 255) << 16;
ctx2.w0[1] = uint_to_hex_lower8 ((a >> 8) & 255) << 0 ctx2.w0[1] = uint_to_hex_lower8 ((a >> 8) & 255) << 0
| uint_to_hex_lower8 ((a >> 0) & 255) << 16; | uint_to_hex_lower8 ((a >> 0) & 255) << 16;
ctx2.w0[2] = uint_to_hex_lower8 ((b >> 24) & 255) << 0 ctx2.w0[2] = uint_to_hex_lower8 ((b >> 24) & 255) << 0
| uint_to_hex_lower8 ((b >> 16) & 255) << 16; | uint_to_hex_lower8 ((b >> 16) & 255) << 16;
ctx2.w0[3] = uint_to_hex_lower8 ((b >> 8) & 255) << 0 ctx2.w0[3] = uint_to_hex_lower8 ((b >> 8) & 255) << 0
| uint_to_hex_lower8 ((b >> 0) & 255) << 16; | uint_to_hex_lower8 ((b >> 0) & 255) << 16;
ctx2.w1[0] = uint_to_hex_lower8 ((c >> 24) & 255) << 0 ctx2.w1[0] = uint_to_hex_lower8 ((c >> 24) & 255) << 0
| uint_to_hex_lower8 ((c >> 16) & 255) << 16; | uint_to_hex_lower8 ((c >> 16) & 255) << 16;
ctx2.w1[1] = uint_to_hex_lower8 ((c >> 8) & 255) << 0 ctx2.w1[1] = uint_to_hex_lower8 ((c >> 8) & 255) << 0
| uint_to_hex_lower8 ((c >> 0) & 255) << 16; | uint_to_hex_lower8 ((c >> 0) & 255) << 16;
ctx2.w1[2] = uint_to_hex_lower8 ((d >> 24) & 255) << 0 ctx2.w1[2] = uint_to_hex_lower8 ((d >> 24) & 255) << 0
| uint_to_hex_lower8 ((d >> 16) & 255) << 16; | uint_to_hex_lower8 ((d >> 16) & 255) << 16;
ctx2.w1[3] = uint_to_hex_lower8 ((d >> 8) & 255) << 0 ctx2.w1[3] = uint_to_hex_lower8 ((d >> 8) & 255) << 0
| uint_to_hex_lower8 ((d >> 0) & 255) << 16; | uint_to_hex_lower8 ((d >> 0) & 255) << 16;
ctx2.w2[0] = uint_to_hex_lower8 ((e >> 24) & 255) << 0 ctx2.w2[0] = uint_to_hex_lower8 ((e >> 24) & 255) << 0
| uint_to_hex_lower8 ((e >> 16) & 255) << 16; | uint_to_hex_lower8 ((e >> 16) & 255) << 16;
ctx2.w2[1] = uint_to_hex_lower8 ((e >> 8) & 255) << 0 ctx2.w2[1] = uint_to_hex_lower8 ((e >> 8) & 255) << 0
| uint_to_hex_lower8 ((e >> 0) & 255) << 16; | uint_to_hex_lower8 ((e >> 0) & 255) << 16;
ctx2.len = 40; ctx2.len = 40;
md5_final (&ctx2); md5_final (&ctx2);
const u32 r0 = ctx2.h[DGST_R0]; const u32 r0 = ctx2.h[DGST_R0];
const u32 r1 = ctx2.h[DGST_R1]; const u32 r1 = ctx2.h[DGST_R1];
const u32 r2 = ctx2.h[DGST_R2]; const u32 r2 = ctx2.h[DGST_R2];
const u32 r3 = ctx2.h[DGST_R3]; const u32 r3 = ctx2.h[DGST_R3];
COMPARE_S_SCALAR (r0, r1, r2, r3); COMPARE_S_SCALAR (r0, r1, r2, r3);
} }
} }

1956
OpenCL/m32800_a1-optimized.cl
View File

File diff suppressed because it is too large Load Diff

602
OpenCL/m32800_a1-pure.cl
View File

@ -1,301 +1,301 @@
/** /**
* Author......: See docs/credits.txt * Author......: See docs/credits.txt
* License.....: MIT * License.....: MIT
*/ */
//#define NEW_SIMD_CODE //#define NEW_SIMD_CODE
#ifdef KERNEL_STATIC #ifdef KERNEL_STATIC
#include M2S(INCLUDE_PATH/inc_vendor.h) #include M2S(INCLUDE_PATH/inc_vendor.h)
#include M2S(INCLUDE_PATH/inc_types.h) #include M2S(INCLUDE_PATH/inc_types.h)
#include M2S(INCLUDE_PATH/inc_platform.cl) #include M2S(INCLUDE_PATH/inc_platform.cl)
#include M2S(INCLUDE_PATH/inc_common.cl) #include M2S(INCLUDE_PATH/inc_common.cl)
#include M2S(INCLUDE_PATH/inc_scalar.cl) #include M2S(INCLUDE_PATH/inc_scalar.cl)
#include M2S(INCLUDE_PATH/inc_hash_md5.cl) #include M2S(INCLUDE_PATH/inc_hash_md5.cl)
#include M2S(INCLUDE_PATH/inc_hash_sha1.cl) #include M2S(INCLUDE_PATH/inc_hash_sha1.cl)
#endif #endif
#if VECT_SIZE == 1 #if VECT_SIZE == 1
#define uint_to_hex_lower8(i) make_u32x (b_bin2asc[(i)]) #define uint_to_hex_lower8(i) make_u32x (b_bin2asc[(i)])
#define uint_to_hex_lower8_le(i) make_u32x (l_bin2asc[(i)]) #define uint_to_hex_lower8_le(i) make_u32x (l_bin2asc[(i)])
#elif VECT_SIZE == 2 #elif VECT_SIZE == 2
#define uint_to_hex_lower8(i) make_u32x (b_bin2asc[(i).s0], b_bin2asc[(i).s1]) #define uint_to_hex_lower8(i) make_u32x (b_bin2asc[(i).s0], b_bin2asc[(i).s1])
#define uint_to_hex_lower8_le(i) make_u32x (l_bin2asc[(i).s0], l_bin2asc[(i).s1]) #define uint_to_hex_lower8_le(i) make_u32x (l_bin2asc[(i).s0], l_bin2asc[(i).s1])
#elif VECT_SIZE == 4 #elif VECT_SIZE == 4
#define uint_to_hex_lower8(i) make_u32x (b_bin2asc[(i).s0], b_bin2asc[(i).s1], b_bin2asc[(i).s2], b_bin2asc[(i).s3]) #define uint_to_hex_lower8(i) make_u32x (b_bin2asc[(i).s0], b_bin2asc[(i).s1], b_bin2asc[(i).s2], b_bin2asc[(i).s3])
#define uint_to_hex_lower8_le(i) make_u32x (l_bin2asc[(i).s0], l_bin2asc[(i).s1], l_bin2asc[(i).s2], l_bin2asc[(i).s3]) #define uint_to_hex_lower8_le(i) make_u32x (l_bin2asc[(i).s0], l_bin2asc[(i).s1], l_bin2asc[(i).s2], l_bin2asc[(i).s3])
#elif VECT_SIZE == 8 #elif VECT_SIZE == 8
#define uint_to_hex_lower8(i) make_u32x (b_bin2asc[(i).s0], b_bin2asc[(i).s1], b_bin2asc[(i).s2], b_bin2asc[(i).s3], b_bin2asc[(i).s4], b_bin2asc[(i).s5], b_bin2asc[(i).s6], b_bin2asc[(i).s7]) #define uint_to_hex_lower8(i) make_u32x (b_bin2asc[(i).s0], b_bin2asc[(i).s1], b_bin2asc[(i).s2], b_bin2asc[(i).s3], b_bin2asc[(i).s4], b_bin2asc[(i).s5], b_bin2asc[(i).s6], b_bin2asc[(i).s7])
#define uint_to_hex_lower8_le(i) make_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]) #define uint_to_hex_lower8_le(i) make_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 #elif VECT_SIZE == 16
#define uint_to_hex_lower8(i) make_u32x (b_bin2asc[(i).s0], b_bin2asc[(i).s1], b_bin2asc[(i).s2], b_bin2asc[(i).s3], b_bin2asc[(i).s4], b_bin2asc[(i).s5], b_bin2asc[(i).s6], b_bin2asc[(i).s7], b_bin2asc[(i).s8], b_bin2asc[(i).s9], b_bin2asc[(i).sa], b_bin2asc[(i).sb], b_bin2asc[(i).sc], b_bin2asc[(i).sd], b_bin2asc[(i).se], b_bin2asc[(i).sf]) #define uint_to_hex_lower8(i) make_u32x (b_bin2asc[(i).s0], b_bin2asc[(i).s1], b_bin2asc[(i).s2], b_bin2asc[(i).s3], b_bin2asc[(i).s4], b_bin2asc[(i).s5], b_bin2asc[(i).s6], b_bin2asc[(i).s7], b_bin2asc[(i).s8], b_bin2asc[(i).s9], b_bin2asc[(i).sa], b_bin2asc[(i).sb], b_bin2asc[(i).sc], b_bin2asc[(i).sd], b_bin2asc[(i).se], b_bin2asc[(i).sf])
#define uint_to_hex_lower8_le(i) make_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]) #define uint_to_hex_lower8_le(i) make_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 #endif
KERNEL_FQ void m32800_mxx (KERN_ATTR_BASIC ()) KERNEL_FQ void m32800_mxx (KERN_ATTR_BASIC ())
{ {
/** /**
* modifier * modifier
*/ */
const u64 gid = get_global_id (0); const u64 gid = get_global_id (0);
const u64 lid = get_local_id (0); const u64 lid = get_local_id (0);
const u64 lsz = get_local_size (0); const u64 lsz = get_local_size (0);
/** /**
* bin2asc tables * bin2asc tables
*/ */
LOCAL_VK u32 l_bin2asc[256]; LOCAL_VK u32 l_bin2asc[256];
LOCAL_VK u32 b_bin2asc[256]; LOCAL_VK u32 b_bin2asc[256];
for (u32 i = lid; i < 256; i += lsz) for (u32 i = lid; i < 256; i += lsz)
{ {
const u32 i0 = (i >> 0) & 15; const u32 i0 = (i >> 0) & 15;
const u32 i1 = (i >> 4) & 15; const u32 i1 = (i >> 4) & 15;
l_bin2asc[i] = ((i0 < 10) ? '0' + i0 : 'a' - 10 + i0) << 0 l_bin2asc[i] = ((i0 < 10) ? '0' + i0 : 'a' - 10 + i0) << 0
| ((i1 < 10) ? '0' + i1 : 'a' - 10 + i1) << 8; | ((i1 < 10) ? '0' + i1 : 'a' - 10 + i1) << 8;
b_bin2asc[i] = ((i0 < 10) ? '0' + i0 : 'a' - 10 + i0) << 8 b_bin2asc[i] = ((i0 < 10) ? '0' + i0 : 'a' - 10 + i0) << 8
| ((i1 < 10) ? '0' + i1 : 'a' - 10 + i1) << 0; | ((i1 < 10) ? '0' + i1 : 'a' - 10 + i1) << 0;
} }
SYNC_THREADS (); SYNC_THREADS ();
if (gid >= GID_CNT) return; if (gid >= GID_CNT) return;
/** /**
* base * base
*/ */
md5_ctx_t ctx; md5_ctx_t ctx;
md5_init (&ctx); md5_init (&ctx);
md5_update_global (&ctx, pws[gid].i, pws[gid].pw_len); md5_update_global (&ctx, pws[gid].i, pws[gid].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++)
{ {
md5_ctx_t ctx0 = ctx; md5_ctx_t ctx0 = ctx;
md5_update_global (&ctx0, combs_buf[il_pos].i, combs_buf[il_pos].pw_len); md5_update_global (&ctx0, combs_buf[il_pos].i, combs_buf[il_pos].pw_len);
md5_final (&ctx0); md5_final (&ctx0);
u32 a = ctx0.h[0]; u32 a = ctx0.h[0];
u32 b = ctx0.h[1]; u32 b = ctx0.h[1];
u32 c = ctx0.h[2]; u32 c = ctx0.h[2];
u32 d = ctx0.h[3]; u32 d = ctx0.h[3];
sha1_ctx_t ctx1; sha1_ctx_t ctx1;
sha1_init (&ctx1); sha1_init (&ctx1);
ctx1.w0[0] = uint_to_hex_lower8_le ((a >> 8) & 255) << 0 ctx1.w0[0] = uint_to_hex_lower8_le ((a >> 8) & 255) << 0
| uint_to_hex_lower8_le ((a >> 0) & 255) << 16; | uint_to_hex_lower8_le ((a >> 0) & 255) << 16;
ctx1.w0[1] = uint_to_hex_lower8_le ((a >> 24) & 255) << 0 ctx1.w0[1] = uint_to_hex_lower8_le ((a >> 24) & 255) << 0
| uint_to_hex_lower8_le ((a >> 16) & 255) << 16; | uint_to_hex_lower8_le ((a >> 16) & 255) << 16;
ctx1.w0[2] = uint_to_hex_lower8_le ((b >> 8) & 255) << 0 ctx1.w0[2] = uint_to_hex_lower8_le ((b >> 8) & 255) << 0
| uint_to_hex_lower8_le ((b >> 0) & 255) << 16; | uint_to_hex_lower8_le ((b >> 0) & 255) << 16;
ctx1.w0[3] = uint_to_hex_lower8_le ((b >> 24) & 255) << 0 ctx1.w0[3] = uint_to_hex_lower8_le ((b >> 24) & 255) << 0
| uint_to_hex_lower8_le ((b >> 16) & 255) << 16; | uint_to_hex_lower8_le ((b >> 16) & 255) << 16;
ctx1.w1[0] = uint_to_hex_lower8_le ((c >> 8) & 255) << 0 ctx1.w1[0] = uint_to_hex_lower8_le ((c >> 8) & 255) << 0
| uint_to_hex_lower8_le ((c >> 0) & 255) << 16; | uint_to_hex_lower8_le ((c >> 0) & 255) << 16;
ctx1.w1[1] = uint_to_hex_lower8_le ((c >> 24) & 255) << 0 ctx1.w1[1] = uint_to_hex_lower8_le ((c >> 24) & 255) << 0
| uint_to_hex_lower8_le ((c >> 16) & 255) << 16; | uint_to_hex_lower8_le ((c >> 16) & 255) << 16;
ctx1.w1[2] = uint_to_hex_lower8_le ((d >> 8) & 255) << 0 ctx1.w1[2] = uint_to_hex_lower8_le ((d >> 8) & 255) << 0
| uint_to_hex_lower8_le ((d >> 0) & 255) << 16; | uint_to_hex_lower8_le ((d >> 0) & 255) << 16;
ctx1.w1[3] = uint_to_hex_lower8_le ((d >> 24) & 255) << 0 ctx1.w1[3] = uint_to_hex_lower8_le ((d >> 24) & 255) << 0
| uint_to_hex_lower8_le ((d >> 16) & 255) << 16; | uint_to_hex_lower8_le ((d >> 16) & 255) << 16;
ctx1.len = 32; ctx1.len = 32;
sha1_final (&ctx1); sha1_final (&ctx1);
a = ctx1.h[0]; a = ctx1.h[0];
b = ctx1.h[1]; b = ctx1.h[1];
c = ctx1.h[2]; c = ctx1.h[2];
d = ctx1.h[3]; d = ctx1.h[3];
const u32 e = ctx1.h[4]; const u32 e = ctx1.h[4];
md5_ctx_t ctx2; md5_ctx_t ctx2;
md5_init (&ctx2); md5_init (&ctx2);
ctx2.w0[0] = uint_to_hex_lower8 ((a >> 24) & 255) << 0 ctx2.w0[0] = uint_to_hex_lower8 ((a >> 24) & 255) << 0
| uint_to_hex_lower8 ((a >> 16) & 255) << 16; | uint_to_hex_lower8 ((a >> 16) & 255) << 16;
ctx2.w0[1] = uint_to_hex_lower8 ((a >> 8) & 255) << 0 ctx2.w0[1] = uint_to_hex_lower8 ((a >> 8) & 255) << 0
| uint_to_hex_lower8 ((a >> 0) & 255) << 16; | uint_to_hex_lower8 ((a >> 0) & 255) << 16;
ctx2.w0[2] = uint_to_hex_lower8 ((b >> 24) & 255) << 0 ctx2.w0[2] = uint_to_hex_lower8 ((b >> 24) & 255) << 0
| uint_to_hex_lower8 ((b >> 16) & 255) << 16; | uint_to_hex_lower8 ((b >> 16) & 255) << 16;
ctx2.w0[3] = uint_to_hex_lower8 ((b >> 8) & 255) << 0 ctx2.w0[3] = uint_to_hex_lower8 ((b >> 8) & 255) << 0
| uint_to_hex_lower8 ((b >> 0) & 255) << 16; | uint_to_hex_lower8 ((b >> 0) & 255) << 16;
ctx2.w1[0] = uint_to_hex_lower8 ((c >> 24) & 255) << 0 ctx2.w1[0] = uint_to_hex_lower8 ((c >> 24) & 255) << 0
| uint_to_hex_lower8 ((c >> 16) & 255) << 16; | uint_to_hex_lower8 ((c >> 16) & 255) << 16;
ctx2.w1[1] = uint_to_hex_lower8 ((c >> 8) & 255) << 0 ctx2.w1[1] = uint_to_hex_lower8 ((c >> 8) & 255) << 0
| uint_to_hex_lower8 ((c >> 0) & 255) << 16; | uint_to_hex_lower8 ((c >> 0) & 255) << 16;
ctx2.w1[2] = uint_to_hex_lower8 ((d >> 24) & 255) << 0 ctx2.w1[2] = uint_to_hex_lower8 ((d >> 24) & 255) << 0
| uint_to_hex_lower8 ((d >> 16) & 255) << 16; | uint_to_hex_lower8 ((d >> 16) & 255) << 16;
ctx2.w1[3] = uint_to_hex_lower8 ((d >> 8) & 255) << 0 ctx2.w1[3] = uint_to_hex_lower8 ((d >> 8) & 255) << 0
| uint_to_hex_lower8 ((d >> 0) & 255) << 16; | uint_to_hex_lower8 ((d >> 0) & 255) << 16;
ctx2.w2[0] = uint_to_hex_lower8 ((e >> 24) & 255) << 0 ctx2.w2[0] = uint_to_hex_lower8 ((e >> 24) & 255) << 0
| uint_to_hex_lower8 ((e >> 16) & 255) << 16; | uint_to_hex_lower8 ((e >> 16) & 255) << 16;
ctx2.w2[1] = uint_to_hex_lower8 ((e >> 8) & 255) << 0 ctx2.w2[1] = uint_to_hex_lower8 ((e >> 8) & 255) << 0
| uint_to_hex_lower8 ((e >> 0) & 255) << 16; | uint_to_hex_lower8 ((e >> 0) & 255) << 16;
ctx2.len = 40; ctx2.len = 40;
md5_final (&ctx2); md5_final (&ctx2);
const u32 r0 = ctx2.h[DGST_R0]; const u32 r0 = ctx2.h[DGST_R0];
const u32 r1 = ctx2.h[DGST_R1]; const u32 r1 = ctx2.h[DGST_R1];
const u32 r2 = ctx2.h[DGST_R2]; const u32 r2 = ctx2.h[DGST_R2];
const u32 r3 = ctx2.h[DGST_R3]; const u32 r3 = ctx2.h[DGST_R3];
COMPARE_M_SCALAR (r0, r1, r2, r3); COMPARE_M_SCALAR (r0, r1, r2, r3);
} }
} }
KERNEL_FQ void m32800_sxx (KERN_ATTR_BASIC ()) KERNEL_FQ void m32800_sxx (KERN_ATTR_BASIC ())
{ {
/** /**
* modifier * modifier
*/ */
const u64 gid = get_global_id (0); const u64 gid = get_global_id (0);
const u64 lid = get_local_id (0); const u64 lid = get_local_id (0);
const u64 lsz = get_local_size (0); const u64 lsz = get_local_size (0);
/** /**
* bin2asc tables * bin2asc tables
*/ */
LOCAL_VK u32 l_bin2asc[256]; LOCAL_VK u32 l_bin2asc[256];
LOCAL_VK u32 b_bin2asc[256]; LOCAL_VK u32 b_bin2asc[256];
for (u32 i = lid; i < 256; i += lsz) for (u32 i = lid; i < 256; i += lsz)
{ {
const u32 i0 = (i >> 0) & 15; const u32 i0 = (i >> 0) & 15;
const u32 i1 = (i >> 4) & 15; const u32 i1 = (i >> 4) & 15;
l_bin2asc[i] = ((i0 < 10) ? '0' + i0 : 'a' - 10 + i0) << 0 l_bin2asc[i] = ((i0 < 10) ? '0' + i0 : 'a' - 10 + i0) << 0
| ((i1 < 10) ? '0' + i1 : 'a' - 10 + i1) << 8; | ((i1 < 10) ? '0' + i1 : 'a' - 10 + i1) << 8;
b_bin2asc[i] = ((i0 < 10) ? '0' + i0 : 'a' - 10 + i0) << 8 b_bin2asc[i] = ((i0 < 10) ? '0' + i0 : 'a' - 10 + i0) << 8
| ((i1 < 10) ? '0' + i1 : 'a' - 10 + i1) << 0; | ((i1 < 10) ? '0' + i1 : 'a' - 10 + i1) << 0;
} }
SYNC_THREADS (); SYNC_THREADS ();
if (gid >= GID_CNT) return; if (gid >= GID_CNT) return;
/** /**
* digest * digest
*/ */
const u32 search[4] = const u32 search[4] =
{ {
digests_buf[DIGESTS_OFFSET_HOST].digest_buf[DGST_R0], 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_R1],
digests_buf[DIGESTS_OFFSET_HOST].digest_buf[DGST_R2], digests_buf[DIGESTS_OFFSET_HOST].digest_buf[DGST_R2],
digests_buf[DIGESTS_OFFSET_HOST].digest_buf[DGST_R3] digests_buf[DIGESTS_OFFSET_HOST].digest_buf[DGST_R3]
}; };
/** /**
* base * base
*/ */
md5_ctx_t ctx; md5_ctx_t ctx;
md5_init (&ctx); md5_init (&ctx);
md5_update_global (&ctx, pws[gid].i, pws[gid].pw_len); md5_update_global (&ctx, pws[gid].i, pws[gid].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++)
{ {
md5_ctx_t ctx0 = ctx; md5_ctx_t ctx0 = ctx;
md5_update_global (&ctx0, combs_buf[il_pos].i, combs_buf[il_pos].pw_len); md5_update_global (&ctx0, combs_buf[il_pos].i, combs_buf[il_pos].pw_len);
md5_final (&ctx0); md5_final (&ctx0);
u32 a = ctx0.h[0]; u32 a = ctx0.h[0];
u32 b = ctx0.h[1]; u32 b = ctx0.h[1];
u32 c = ctx0.h[2]; u32 c = ctx0.h[2];
u32 d = ctx0.h[3]; u32 d = ctx0.h[3];
sha1_ctx_t ctx1; sha1_ctx_t ctx1;
sha1_init (&ctx1); sha1_init (&ctx1);
ctx1.w0[0] = uint_to_hex_lower8_le ((a >> 8) & 255) << 0 ctx1.w0[0] = uint_to_hex_lower8_le ((a >> 8) & 255) << 0
| uint_to_hex_lower8_le ((a >> 0) & 255) << 16; | uint_to_hex_lower8_le ((a >> 0) & 255) << 16;
ctx1.w0[1] = uint_to_hex_lower8_le ((a >> 24) & 255) << 0 ctx1.w0[1] = uint_to_hex_lower8_le ((a >> 24) & 255) << 0
| uint_to_hex_lower8_le ((a >> 16) & 255) << 16; | uint_to_hex_lower8_le ((a >> 16) & 255) << 16;
ctx1.w0[2] = uint_to_hex_lower8_le ((b >> 8) & 255) << 0 ctx1.w0[2] = uint_to_hex_lower8_le ((b >> 8) & 255) << 0
| uint_to_hex_lower8_le ((b >> 0) & 255) << 16; | uint_to_hex_lower8_le ((b >> 0) & 255) << 16;
ctx1.w0[3] = uint_to_hex_lower8_le ((b >> 24) & 255) << 0 ctx1.w0[3] = uint_to_hex_lower8_le ((b >> 24) & 255) << 0
| uint_to_hex_lower8_le ((b >> 16) & 255) << 16; | uint_to_hex_lower8_le ((b >> 16) & 255) << 16;
ctx1.w1[0] = uint_to_hex_lower8_le ((c >> 8) & 255) << 0 ctx1.w1[0] = uint_to_hex_lower8_le ((c >> 8) & 255) << 0
| uint_to_hex_lower8_le ((c >> 0) & 255) << 16; | uint_to_hex_lower8_le ((c >> 0) & 255) << 16;
ctx1.w1[1] = uint_to_hex_lower8_le ((c >> 24) & 255) << 0 ctx1.w1[1] = uint_to_hex_lower8_le ((c >> 24) & 255) << 0
| uint_to_hex_lower8_le ((c >> 16) & 255) << 16; | uint_to_hex_lower8_le ((c >> 16) & 255) << 16;
ctx1.w1[2] = uint_to_hex_lower8_le ((d >> 8) & 255) << 0 ctx1.w1[2] = uint_to_hex_lower8_le ((d >> 8) & 255) << 0
| uint_to_hex_lower8_le ((d >> 0) & 255) << 16; | uint_to_hex_lower8_le ((d >> 0) & 255) << 16;
ctx1.w1[3] = uint_to_hex_lower8_le ((d >> 24) & 255) << 0 ctx1.w1[3] = uint_to_hex_lower8_le ((d >> 24) & 255) << 0
| uint_to_hex_lower8_le ((d >> 16) & 255) << 16; | uint_to_hex_lower8_le ((d >> 16) & 255) << 16;
ctx1.len = 32; ctx1.len = 32;
sha1_final (&ctx1); sha1_final (&ctx1);
a = ctx1.h[0]; a = ctx1.h[0];
b = ctx1.h[1]; b = ctx1.h[1];
c = ctx1.h[2]; c = ctx1.h[2];
d = ctx1.h[3]; d = ctx1.h[3];
const u32 e = ctx1.h[4]; const u32 e = ctx1.h[4];
md5_ctx_t ctx2; md5_ctx_t ctx2;
md5_init (&ctx2); md5_init (&ctx2);
ctx2.w0[0] = uint_to_hex_lower8 ((a >> 24) & 255) << 0 ctx2.w0[0] = uint_to_hex_lower8 ((a >> 24) & 255) << 0
| uint_to_hex_lower8 ((a >> 16) & 255) << 16; | uint_to_hex_lower8 ((a >> 16) & 255) << 16;
ctx2.w0[1] = uint_to_hex_lower8 ((a >> 8) & 255) << 0 ctx2.w0[1] = uint_to_hex_lower8 ((a >> 8) & 255) << 0
| uint_to_hex_lower8 ((a >> 0) & 255) << 16; | uint_to_hex_lower8 ((a >> 0) & 255) << 16;
ctx2.w0[2] = uint_to_hex_lower8 ((b >> 24) & 255) << 0 ctx2.w0[2] = uint_to_hex_lower8 ((b >> 24) & 255) << 0
| uint_to_hex_lower8 ((b >> 16) & 255) << 16; | uint_to_hex_lower8 ((b >> 16) & 255) << 16;
ctx2.w0[3] = uint_to_hex_lower8 ((b >> 8) & 255) << 0 ctx2.w0[3] = uint_to_hex_lower8 ((b >> 8) & 255) << 0
| uint_to_hex_lower8 ((b >> 0) & 255) << 16; | uint_to_hex_lower8 ((b >> 0) & 255) << 16;
ctx2.w1[0] = uint_to_hex_lower8 ((c >> 24) & 255) << 0 ctx2.w1[0] = uint_to_hex_lower8 ((c >> 24) & 255) << 0
| uint_to_hex_lower8 ((c >> 16) & 255) << 16; | uint_to_hex_lower8 ((c >> 16) & 255) << 16;
ctx2.w1[1] = uint_to_hex_lower8 ((c >> 8) & 255) << 0 ctx2.w1[1] = uint_to_hex_lower8 ((c >> 8) & 255) << 0
| uint_to_hex_lower8 ((c >> 0) & 255) << 16; | uint_to_hex_lower8 ((c >> 0) & 255) << 16;
ctx2.w1[2] = uint_to_hex_lower8 ((d >> 24) & 255) << 0 ctx2.w1[2] = uint_to_hex_lower8 ((d >> 24) & 255) << 0
| uint_to_hex_lower8 ((d >> 16) & 255) << 16; | uint_to_hex_lower8 ((d >> 16) & 255) << 16;
ctx2.w1[3] = uint_to_hex_lower8 ((d >> 8) & 255) << 0 ctx2.w1[3] = uint_to_hex_lower8 ((d >> 8) & 255) << 0
| uint_to_hex_lower8 ((d >> 0) & 255) << 16; | uint_to_hex_lower8 ((d >> 0) & 255) << 16;
ctx2.w2[0] = uint_to_hex_lower8 ((e >> 24) & 255) << 0 ctx2.w2[0] = uint_to_hex_lower8 ((e >> 24) & 255) << 0
| uint_to_hex_lower8 ((e >> 16) & 255) << 16; | uint_to_hex_lower8 ((e >> 16) & 255) << 16;
ctx2.w2[1] = uint_to_hex_lower8 ((e >> 8) & 255) << 0 ctx2.w2[1] = uint_to_hex_lower8 ((e >> 8) & 255) << 0
| uint_to_hex_lower8 ((e >> 0) & 255) << 16; | uint_to_hex_lower8 ((e >> 0) & 255) << 16;
ctx2.len = 40; ctx2.len = 40;
md5_final (&ctx2); md5_final (&ctx2);
const u32 r0 = ctx2.h[DGST_R0]; const u32 r0 = ctx2.h[DGST_R0];
const u32 r1 = ctx2.h[DGST_R1]; const u32 r1 = ctx2.h[DGST_R1];
const u32 r2 = ctx2.h[DGST_R2]; const u32 r2 = ctx2.h[DGST_R2];
const u32 r3 = ctx2.h[DGST_R3]; const u32 r3 = ctx2.h[DGST_R3];
COMPARE_S_SCALAR (r0, r1, r2, r3); COMPARE_S_SCALAR (r0, r1, r2, r3);
} }
} }

2444
OpenCL/m32800_a3-optimized.cl
View File

File diff suppressed because it is too large Load Diff

654
OpenCL/m32800_a3-pure.cl
View File

@ -1,327 +1,327 @@
/** /**
* Author......: See docs/credits.txt * Author......: See docs/credits.txt
* License.....: MIT * License.....: MIT
*/ */
#define NEW_SIMD_CODE #define NEW_SIMD_CODE
#ifdef KERNEL_STATIC #ifdef KERNEL_STATIC
#include M2S(INCLUDE_PATH/inc_vendor.h) #include M2S(INCLUDE_PATH/inc_vendor.h)
#include M2S(INCLUDE_PATH/inc_types.h) #include M2S(INCLUDE_PATH/inc_types.h)
#include M2S(INCLUDE_PATH/inc_platform.cl) #include M2S(INCLUDE_PATH/inc_platform.cl)
#include M2S(INCLUDE_PATH/inc_common.cl) #include M2S(INCLUDE_PATH/inc_common.cl)
#include M2S(INCLUDE_PATH/inc_simd.cl) #include M2S(INCLUDE_PATH/inc_simd.cl)
#include M2S(INCLUDE_PATH/inc_hash_md5.cl) #include M2S(INCLUDE_PATH/inc_hash_md5.cl)
#include M2S(INCLUDE_PATH/inc_hash_sha1.cl) #include M2S(INCLUDE_PATH/inc_hash_sha1.cl)
#endif #endif
#if VECT_SIZE == 1 #if VECT_SIZE == 1
#define uint_to_hex_lower8(i) make_u32x (b_bin2asc[(i)]) #define uint_to_hex_lower8(i) make_u32x (b_bin2asc[(i)])
#define uint_to_hex_lower8_le(i) make_u32x (l_bin2asc[(i)]) #define uint_to_hex_lower8_le(i) make_u32x (l_bin2asc[(i)])
#elif VECT_SIZE == 2 #elif VECT_SIZE == 2
#define uint_to_hex_lower8(i) make_u32x (b_bin2asc[(i).s0], b_bin2asc[(i).s1]) #define uint_to_hex_lower8(i) make_u32x (b_bin2asc[(i).s0], b_bin2asc[(i).s1])
#define uint_to_hex_lower8_le(i) make_u32x (l_bin2asc[(i).s0], l_bin2asc[(i).s1]) #define uint_to_hex_lower8_le(i) make_u32x (l_bin2asc[(i).s0], l_bin2asc[(i).s1])
#elif VECT_SIZE == 4 #elif VECT_SIZE == 4
#define uint_to_hex_lower8(i) make_u32x (b_bin2asc[(i).s0], b_bin2asc[(i).s1], b_bin2asc[(i).s2], b_bin2asc[(i).s3]) #define uint_to_hex_lower8(i) make_u32x (b_bin2asc[(i).s0], b_bin2asc[(i).s1], b_bin2asc[(i).s2], b_bin2asc[(i).s3])
#define uint_to_hex_lower8_le(i) make_u32x (l_bin2asc[(i).s0], l_bin2asc[(i).s1], l_bin2asc[(i).s2], l_bin2asc[(i).s3]) #define uint_to_hex_lower8_le(i) make_u32x (l_bin2asc[(i).s0], l_bin2asc[(i).s1], l_bin2asc[(i).s2], l_bin2asc[(i).s3])
#elif VECT_SIZE == 8 #elif VECT_SIZE == 8
#define uint_to_hex_lower8(i) make_u32x (b_bin2asc[(i).s0], b_bin2asc[(i).s1], b_bin2asc[(i).s2], b_bin2asc[(i).s3], b_bin2asc[(i).s4], b_bin2asc[(i).s5], b_bin2asc[(i).s6], b_bin2asc[(i).s7]) #define uint_to_hex_lower8(i) make_u32x (b_bin2asc[(i).s0], b_bin2asc[(i).s1], b_bin2asc[(i).s2], b_bin2asc[(i).s3], b_bin2asc[(i).s4], b_bin2asc[(i).s5], b_bin2asc[(i).s6], b_bin2asc[(i).s7])
#define uint_to_hex_lower8_le(i) make_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]) #define uint_to_hex_lower8_le(i) make_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 #elif VECT_SIZE == 16
#define uint_to_hex_lower8(i) make_u32x (b_bin2asc[(i).s0], b_bin2asc[(i).s1], b_bin2asc[(i).s2], b_bin2asc[(i).s3], b_bin2asc[(i).s4], b_bin2asc[(i).s5], b_bin2asc[(i).s6], b_bin2asc[(i).s7], b_bin2asc[(i).s8], b_bin2asc[(i).s9], b_bin2asc[(i).sa], b_bin2asc[(i).sb], b_bin2asc[(i).sc], b_bin2asc[(i).sd], b_bin2asc[(i).se], b_bin2asc[(i).sf]) #define uint_to_hex_lower8(i) make_u32x (b_bin2asc[(i).s0], b_bin2asc[(i).s1], b_bin2asc[(i).s2], b_bin2asc[(i).s3], b_bin2asc[(i).s4], b_bin2asc[(i).s5], b_bin2asc[(i).s6], b_bin2asc[(i).s7], b_bin2asc[(i).s8], b_bin2asc[(i).s9], b_bin2asc[(i).sa], b_bin2asc[(i).sb], b_bin2asc[(i).sc], b_bin2asc[(i).sd], b_bin2asc[(i).se], b_bin2asc[(i).sf])
#define uint_to_hex_lower8_le(i) make_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]) #define uint_to_hex_lower8_le(i) make_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 #endif
KERNEL_FQ void m32800_mxx (KERN_ATTR_VECTOR ()) KERNEL_FQ void m32800_mxx (KERN_ATTR_VECTOR ())
{ {
/** /**
* modifier * modifier
*/ */
const u64 gid = get_global_id (0); const u64 gid = get_global_id (0);
const u64 lid = get_local_id (0); const u64 lid = get_local_id (0);
const u64 lsz = get_local_size (0); const u64 lsz = get_local_size (0);
/** /**
* bin2asc tables * bin2asc tables
*/ */
LOCAL_VK u32 l_bin2asc[256]; LOCAL_VK u32 l_bin2asc[256];
LOCAL_VK u32 b_bin2asc[256]; LOCAL_VK u32 b_bin2asc[256];
for (u32 i = lid; i < 256; i += lsz) for (u32 i = lid; i < 256; i += lsz)
{ {
const u32 i0 = (i >> 0) & 15; const u32 i0 = (i >> 0) & 15;
const u32 i1 = (i >> 4) & 15; const u32 i1 = (i >> 4) & 15;
l_bin2asc[i] = ((i0 < 10) ? '0' + i0 : 'a' - 10 + i0) << 0 l_bin2asc[i] = ((i0 < 10) ? '0' + i0 : 'a' - 10 + i0) << 0
| ((i1 < 10) ? '0' + i1 : 'a' - 10 + i1) << 8; | ((i1 < 10) ? '0' + i1 : 'a' - 10 + i1) << 8;
b_bin2asc[i] = ((i0 < 10) ? '0' + i0 : 'a' - 10 + i0) << 8 b_bin2asc[i] = ((i0 < 10) ? '0' + i0 : 'a' - 10 + i0) << 8
| ((i1 < 10) ? '0' + i1 : 'a' - 10 + i1) << 0; | ((i1 < 10) ? '0' + i1 : 'a' - 10 + i1) << 0;
} }
SYNC_THREADS (); SYNC_THREADS ();
if (gid >= GID_CNT) return; if (gid >= GID_CNT) return;
/** /**
* base * base
*/ */
const u32 pw_len = pws[gid].pw_len; const u32 pw_len = pws[gid].pw_len;
u32x w[64] = { 0 }; u32x w[64] = { 0 };
for (u32 i = 0, idx = 0; i < pw_len; i += 4, idx += 1) for (u32 i = 0, idx = 0; i < pw_len; i += 4, idx += 1)
{ {
w[idx] = pws[gid].i[idx]; w[idx] = pws[gid].i[idx];
} }
/** /**
* loop * loop
*/ */
u32x w0l = w[0]; u32x w0l = w[0];
for (u32 il_pos = 0; il_pos < IL_CNT; il_pos += VECT_SIZE) 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 w0r = words_buf_r[il_pos / VECT_SIZE];
const u32x w0 = w0l | w0r; const u32x w0 = w0l | w0r;
w[0] = w0; w[0] = w0;
md5_ctx_vector_t ctx0; md5_ctx_vector_t ctx0;
md5_init_vector (&ctx0); md5_init_vector (&ctx0);
md5_update_vector (&ctx0, w, pw_len); md5_update_vector (&ctx0, w, pw_len);
md5_final_vector (&ctx0); md5_final_vector (&ctx0);
u32 a = ctx0.h[0]; u32 a = ctx0.h[0];
u32 b = ctx0.h[1]; u32 b = ctx0.h[1];
u32 c = ctx0.h[2]; u32 c = ctx0.h[2];
u32 d = ctx0.h[3]; u32 d = ctx0.h[3];
sha1_ctx_vector_t ctx1; sha1_ctx_vector_t ctx1;
sha1_init_vector (&ctx1); sha1_init_vector (&ctx1);
ctx1.w0[0] = uint_to_hex_lower8_le ((a >> 8) & 255) << 0 ctx1.w0[0] = uint_to_hex_lower8_le ((a >> 8) & 255) << 0
| uint_to_hex_lower8_le ((a >> 0) & 255) << 16; | uint_to_hex_lower8_le ((a >> 0) & 255) << 16;
ctx1.w0[1] = uint_to_hex_lower8_le ((a >> 24) & 255) << 0 ctx1.w0[1] = uint_to_hex_lower8_le ((a >> 24) & 255) << 0
| uint_to_hex_lower8_le ((a >> 16) & 255) << 16; | uint_to_hex_lower8_le ((a >> 16) & 255) << 16;
ctx1.w0[2] = uint_to_hex_lower8_le ((b >> 8) & 255) << 0 ctx1.w0[2] = uint_to_hex_lower8_le ((b >> 8) & 255) << 0
| uint_to_hex_lower8_le ((b >> 0) & 255) << 16; | uint_to_hex_lower8_le ((b >> 0) & 255) << 16;
ctx1.w0[3] = uint_to_hex_lower8_le ((b >> 24) & 255) << 0 ctx1.w0[3] = uint_to_hex_lower8_le ((b >> 24) & 255) << 0
| uint_to_hex_lower8_le ((b >> 16) & 255) << 16; | uint_to_hex_lower8_le ((b >> 16) & 255) << 16;
ctx1.w1[0] = uint_to_hex_lower8_le ((c >> 8) & 255) << 0 ctx1.w1[0] = uint_to_hex_lower8_le ((c >> 8) & 255) << 0
| uint_to_hex_lower8_le ((c >> 0) & 255) << 16; | uint_to_hex_lower8_le ((c >> 0) & 255) << 16;
ctx1.w1[1] = uint_to_hex_lower8_le ((c >> 24) & 255) << 0 ctx1.w1[1] = uint_to_hex_lower8_le ((c >> 24) & 255) << 0
| uint_to_hex_lower8_le ((c >> 16) & 255) << 16; | uint_to_hex_lower8_le ((c >> 16) & 255) << 16;
ctx1.w1[2] = uint_to_hex_lower8_le ((d >> 8) & 255) << 0 ctx1.w1[2] = uint_to_hex_lower8_le ((d >> 8) & 255) << 0
| uint_to_hex_lower8_le ((d >> 0) & 255) << 16; | uint_to_hex_lower8_le ((d >> 0) & 255) << 16;
ctx1.w1[3] = uint_to_hex_lower8_le ((d >> 24) & 255) << 0 ctx1.w1[3] = uint_to_hex_lower8_le ((d >> 24) & 255) << 0
| uint_to_hex_lower8_le ((d >> 16) & 255) << 16; | uint_to_hex_lower8_le ((d >> 16) & 255) << 16;
ctx1.len = 32; ctx1.len = 32;
sha1_final_vector (&ctx1); sha1_final_vector (&ctx1);
a = ctx1.h[0]; a = ctx1.h[0];
b = ctx1.h[1]; b = ctx1.h[1];
c = ctx1.h[2]; c = ctx1.h[2];
d = ctx1.h[3]; d = ctx1.h[3];
const u32 e = ctx1.h[4]; const u32 e = ctx1.h[4];
md5_ctx_vector_t ctx2; md5_ctx_vector_t ctx2;
md5_init_vector (&ctx2); md5_init_vector (&ctx2);
ctx2.w0[0] = uint_to_hex_lower8 ((a >> 24) & 255) << 0 ctx2.w0[0] = uint_to_hex_lower8 ((a >> 24) & 255) << 0
| uint_to_hex_lower8 ((a >> 16) & 255) << 16; | uint_to_hex_lower8 ((a >> 16) & 255) << 16;
ctx2.w0[1] = uint_to_hex_lower8 ((a >> 8) & 255) << 0 ctx2.w0[1] = uint_to_hex_lower8 ((a >> 8) & 255) << 0
| uint_to_hex_lower8 ((a >> 0) & 255) << 16; | uint_to_hex_lower8 ((a >> 0) & 255) << 16;
ctx2.w0[2] = uint_to_hex_lower8 ((b >> 24) & 255) << 0 ctx2.w0[2] = uint_to_hex_lower8 ((b >> 24) & 255) << 0
| uint_to_hex_lower8 ((b >> 16) & 255) << 16; | uint_to_hex_lower8 ((b >> 16) & 255) << 16;
ctx2.w0[3] = uint_to_hex_lower8 ((b >> 8) & 255) << 0 ctx2.w0[3] = uint_to_hex_lower8 ((b >> 8) & 255) << 0
| uint_to_hex_lower8 ((b >> 0) & 255) << 16; | uint_to_hex_lower8 ((b >> 0) & 255) << 16;
ctx2.w1[0] = uint_to_hex_lower8 ((c >> 24) & 255) << 0 ctx2.w1[0] = uint_to_hex_lower8 ((c >> 24) & 255) << 0
| uint_to_hex_lower8 ((c >> 16) & 255) << 16; | uint_to_hex_lower8 ((c >> 16) & 255) << 16;
ctx2.w1[1] = uint_to_hex_lower8 ((c >> 8) & 255) << 0 ctx2.w1[1] = uint_to_hex_lower8 ((c >> 8) & 255) << 0
| uint_to_hex_lower8 ((c >> 0) & 255) << 16; | uint_to_hex_lower8 ((c >> 0) & 255) << 16;
ctx2.w1[2] = uint_to_hex_lower8 ((d >> 24) & 255) << 0 ctx2.w1[2] = uint_to_hex_lower8 ((d >> 24) & 255) << 0
| uint_to_hex_lower8 ((d >> 16) & 255) << 16; | uint_to_hex_lower8 ((d >> 16) & 255) << 16;
ctx2.w1[3] = uint_to_hex_lower8 ((d >> 8) & 255) << 0 ctx2.w1[3] = uint_to_hex_lower8 ((d >> 8) & 255) << 0
| uint_to_hex_lower8 ((d >> 0) & 255) << 16; | uint_to_hex_lower8 ((d >> 0) & 255) << 16;
ctx2.w2[0] = uint_to_hex_lower8 ((e >> 24) & 255) << 0 ctx2.w2[0] = uint_to_hex_lower8 ((e >> 24) & 255) << 0
| uint_to_hex_lower8 ((e >> 16) & 255) << 16; | uint_to_hex_lower8 ((e >> 16) & 255) << 16;
ctx2.w2[1] = uint_to_hex_lower8 ((e >> 8) & 255) << 0 ctx2.w2[1] = uint_to_hex_lower8 ((e >> 8) & 255) << 0
| uint_to_hex_lower8 ((e >> 0) & 255) << 16; | uint_to_hex_lower8 ((e >> 0) & 255) << 16;
ctx2.len = 40; ctx2.len = 40;
md5_final_vector (&ctx2); md5_final_vector (&ctx2);
const u32 r0 = ctx2.h[DGST_R0]; const u32 r0 = ctx2.h[DGST_R0];
const u32 r1 = ctx2.h[DGST_R1]; const u32 r1 = ctx2.h[DGST_R1];
const u32 r2 = ctx2.h[DGST_R2]; const u32 r2 = ctx2.h[DGST_R2];
const u32 r3 = ctx2.h[DGST_R3]; const u32 r3 = ctx2.h[DGST_R3];
COMPARE_M_SIMD (r0, r1, r2, r3); COMPARE_M_SIMD (r0, r1, r2, r3);
} }
} }
KERNEL_FQ void m32800_sxx (KERN_ATTR_VECTOR ()) KERNEL_FQ void m32800_sxx (KERN_ATTR_VECTOR ())
{ {
/** /**
* modifier * modifier
*/ */
const u64 gid = get_global_id (0); const u64 gid = get_global_id (0);
const u64 lid = get_local_id (0); const u64 lid = get_local_id (0);
const u64 lsz = get_local_size (0); const u64 lsz = get_local_size (0);
/** /**
* bin2asc tables * bin2asc tables
*/ */
LOCAL_VK u32 l_bin2asc[256]; LOCAL_VK u32 l_bin2asc[256];
LOCAL_VK u32 b_bin2asc[256]; LOCAL_VK u32 b_bin2asc[256];
for (u32 i = lid; i < 256; i += lsz) for (u32 i = lid; i < 256; i += lsz)
{ {
const u32 i0 = (i >> 0) & 15; const u32 i0 = (i >> 0) & 15;
const u32 i1 = (i >> 4) & 15; const u32 i1 = (i >> 4) & 15;
l_bin2asc[i] = ((i0 < 10) ? '0' + i0 : 'a' - 10 + i0) << 0 l_bin2asc[i] = ((i0 < 10) ? '0' + i0 : 'a' - 10 + i0) << 0
| ((i1 < 10) ? '0' + i1 : 'a' - 10 + i1) << 8; | ((i1 < 10) ? '0' + i1 : 'a' - 10 + i1) << 8;
b_bin2asc[i] = ((i0 < 10) ? '0' + i0 : 'a' - 10 + i0) << 8 b_bin2asc[i] = ((i0 < 10) ? '0' + i0 : 'a' - 10 + i0) << 8
| ((i1 < 10) ? '0' + i1 : 'a' - 10 + i1) << 0; | ((i1 < 10) ? '0' + i1 : 'a' - 10 + i1) << 0;
} }
SYNC_THREADS (); SYNC_THREADS ();
if (gid >= GID_CNT) return; if (gid >= GID_CNT) return;
/** /**
* digest * digest
*/ */
const u32 search[4] = const u32 search[4] =
{ {
digests_buf[DIGESTS_OFFSET_HOST].digest_buf[DGST_R0], 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_R1],
digests_buf[DIGESTS_OFFSET_HOST].digest_buf[DGST_R2], digests_buf[DIGESTS_OFFSET_HOST].digest_buf[DGST_R2],
digests_buf[DIGESTS_OFFSET_HOST].digest_buf[DGST_R3] digests_buf[DIGESTS_OFFSET_HOST].digest_buf[DGST_R3]
}; };
/** /**
* base * base
*/ */
const u32 pw_len = pws[gid].pw_len; const u32 pw_len = pws[gid].pw_len;
u32x w[64] = { 0 }; u32x w[64] = { 0 };
for (u32 i = 0, idx = 0; i < pw_len; i += 4, idx += 1) for (u32 i = 0, idx = 0; i < pw_len; i += 4, idx += 1)
{ {
w[idx] = pws[gid].i[idx]; w[idx] = pws[gid].i[idx];
} }
/** /**
* loop * loop
*/ */
u32x w0l = w[0]; u32x w0l = w[0];
for (u32 il_pos = 0; il_pos < IL_CNT; il_pos += VECT_SIZE) 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 w0r = words_buf_r[il_pos / VECT_SIZE];
const u32x w0 = w0l | w0r; const u32x w0 = w0l | w0r;
w[0] = w0; w[0] = w0;
md5_ctx_vector_t ctx0; md5_ctx_vector_t ctx0;
md5_init_vector (&ctx0); md5_init_vector (&ctx0);
md5_update_vector (&ctx0, w, pw_len); md5_update_vector (&ctx0, w, pw_len);
md5_final_vector (&ctx0); md5_final_vector (&ctx0);
u32 a = ctx0.h[0]; u32 a = ctx0.h[0];
u32 b = ctx0.h[1]; u32 b = ctx0.h[1];
u32 c = ctx0.h[2]; u32 c = ctx0.h[2];
u32 d = ctx0.h[3]; u32 d = ctx0.h[3];
sha1_ctx_vector_t ctx1; sha1_ctx_vector_t ctx1;
sha1_init_vector (&ctx1); sha1_init_vector (&ctx1);
ctx1.w0[0] = uint_to_hex_lower8_le ((a >> 8) & 255) << 0 ctx1.w0[0] = uint_to_hex_lower8_le ((a >> 8) & 255) << 0
| uint_to_hex_lower8_le ((a >> 0) & 255) << 16; | uint_to_hex_lower8_le ((a >> 0) & 255) << 16;
ctx1.w0[1] = uint_to_hex_lower8_le ((a >> 24) & 255) << 0 ctx1.w0[1] = uint_to_hex_lower8_le ((a >> 24) & 255) << 0
| uint_to_hex_lower8_le ((a >> 16) & 255) << 16; | uint_to_hex_lower8_le ((a >> 16) & 255) << 16;
ctx1.w0[2] = uint_to_hex_lower8_le ((b >> 8) & 255) << 0 ctx1.w0[2] = uint_to_hex_lower8_le ((b >> 8) & 255) << 0
| uint_to_hex_lower8_le ((b >> 0) & 255) << 16; | uint_to_hex_lower8_le ((b >> 0) & 255) << 16;
ctx1.w0[3] = uint_to_hex_lower8_le ((b >> 24) & 255) << 0 ctx1.w0[3] = uint_to_hex_lower8_le ((b >> 24) & 255) << 0
| uint_to_hex_lower8_le ((b >> 16) & 255) << 16; | uint_to_hex_lower8_le ((b >> 16) & 255) << 16;
ctx1.w1[0] = uint_to_hex_lower8_le ((c >> 8) & 255) << 0 ctx1.w1[0] = uint_to_hex_lower8_le ((c >> 8) & 255) << 0
| uint_to_hex_lower8_le ((c >> 0) & 255) << 16; | uint_to_hex_lower8_le ((c >> 0) & 255) << 16;
ctx1.w1[1] = uint_to_hex_lower8_le ((c >> 24) & 255) << 0 ctx1.w1[1] = uint_to_hex_lower8_le ((c >> 24) & 255) << 0
| uint_to_hex_lower8_le ((c >> 16) & 255) << 16; | uint_to_hex_lower8_le ((c >> 16) & 255) << 16;
ctx1.w1[2] = uint_to_hex_lower8_le ((d >> 8) & 255) << 0 ctx1.w1[2] = uint_to_hex_lower8_le ((d >> 8) & 255) << 0
| uint_to_hex_lower8_le ((d >> 0) & 255) << 16; | uint_to_hex_lower8_le ((d >> 0) & 255) << 16;
ctx1.w1[3] = uint_to_hex_lower8_le ((d >> 24) & 255) << 0 ctx1.w1[3] = uint_to_hex_lower8_le ((d >> 24) & 255) << 0
| uint_to_hex_lower8_le ((d >> 16) & 255) << 16; | uint_to_hex_lower8_le ((d >> 16) & 255) << 16;
ctx1.len = 32; ctx1.len = 32;
sha1_final_vector (&ctx1); sha1_final_vector (&ctx1);
a = ctx1.h[0]; a = ctx1.h[0];
b = ctx1.h[1]; b = ctx1.h[1];
c = ctx1.h[2]; c = ctx1.h[2];
d = ctx1.h[3]; d = ctx1.h[3];
const u32 e = ctx1.h[4]; const u32 e = ctx1.h[4];
md5_ctx_vector_t ctx2; md5_ctx_vector_t ctx2;
md5_init_vector (&ctx2); md5_init_vector (&ctx2);
ctx2.w0[0] = uint_to_hex_lower8 ((a >> 24) & 255) << 0 ctx2.w0[0] = uint_to_hex_lower8 ((a >> 24) & 255) << 0
| uint_to_hex_lower8 ((a >> 16) & 255) << 16; | uint_to_hex_lower8 ((a >> 16) & 255) << 16;
ctx2.w0[1] = uint_to_hex_lower8 ((a >> 8) & 255) << 0 ctx2.w0[1] = uint_to_hex_lower8 ((a >> 8) & 255) << 0
| uint_to_hex_lower8 ((a >> 0) & 255) << 16; | uint_to_hex_lower8 ((a >> 0) & 255) << 16;
ctx2.w0[2] = uint_to_hex_lower8 ((b >> 24) & 255) << 0 ctx2.w0[2] = uint_to_hex_lower8 ((b >> 24) & 255) << 0
| uint_to_hex_lower8 ((b >> 16) & 255) << 16; | uint_to_hex_lower8 ((b >> 16) & 255) << 16;
ctx2.w0[3] = uint_to_hex_lower8 ((b >> 8) & 255) << 0 ctx2.w0[3] = uint_to_hex_lower8 ((b >> 8) & 255) << 0
| uint_to_hex_lower8 ((b >> 0) & 255) << 16; | uint_to_hex_lower8 ((b >> 0) & 255) << 16;
ctx2.w1[0] = uint_to_hex_lower8 ((c >> 24) & 255) << 0 ctx2.w1[0] = uint_to_hex_lower8 ((c >> 24) & 255) << 0
| uint_to_hex_lower8 ((c >> 16) & 255) << 16; | uint_to_hex_lower8 ((c >> 16) & 255) << 16;
ctx2.w1[1] = uint_to_hex_lower8 ((c >> 8) & 255) << 0 ctx2.w1[1] = uint_to_hex_lower8 ((c >> 8) & 255) << 0
| uint_to_hex_lower8 ((c >> 0) & 255) << 16; | uint_to_hex_lower8 ((c >> 0) & 255) << 16;
ctx2.w1[2] = uint_to_hex_lower8 ((d >> 24) & 255) << 0 ctx2.w1[2] = uint_to_hex_lower8 ((d >> 24) & 255) << 0
| uint_to_hex_lower8 ((d >> 16) & 255) << 16; | uint_to_hex_lower8 ((d >> 16) & 255) << 16;
ctx2.w1[3] = uint_to_hex_lower8 ((d >> 8) & 255) << 0 ctx2.w1[3] = uint_to_hex_lower8 ((d >> 8) & 255) << 0
| uint_to_hex_lower8 ((d >> 0) & 255) << 16; | uint_to_hex_lower8 ((d >> 0) & 255) << 16;
ctx2.w2[0] = uint_to_hex_lower8 ((e >> 24) & 255) << 0 ctx2.w2[0] = uint_to_hex_lower8 ((e >> 24) & 255) << 0
| uint_to_hex_lower8 ((e >> 16) & 255) << 16; | uint_to_hex_lower8 ((e >> 16) & 255) << 16;
ctx2.w2[1] = uint_to_hex_lower8 ((e >> 8) & 255) << 0 ctx2.w2[1] = uint_to_hex_lower8 ((e >> 8) & 255) << 0
| uint_to_hex_lower8 ((e >> 0) & 255) << 16; | uint_to_hex_lower8 ((e >> 0) & 255) << 16;
ctx2.len = 40; ctx2.len = 40;
md5_final_vector (&ctx2); md5_final_vector (&ctx2);
const u32 r0 = ctx2.h[DGST_R0]; const u32 r0 = ctx2.h[DGST_R0];
const u32 r1 = ctx2.h[DGST_R1]; const u32 r1 = ctx2.h[DGST_R1];
const u32 r2 = ctx2.h[DGST_R2]; const u32 r2 = ctx2.h[DGST_R2];
const u32 r3 = ctx2.h[DGST_R3]; const u32 r3 = ctx2.h[DGST_R3];
COMPARE_S_SIMD (r0, r1, r2, r3); COMPARE_S_SIMD (r0, r1, r2, r3);
} }
} }

2
src/modules/module_32800.c
View File

@ -129,6 +129,8 @@ void module_init (module_ctx_t *module_ctx)
module_ctx->module_benchmark_mask = MODULE_DEFAULT; module_ctx->module_benchmark_mask = MODULE_DEFAULT;
module_ctx->module_benchmark_charset = MODULE_DEFAULT; module_ctx->module_benchmark_charset = MODULE_DEFAULT;
module_ctx->module_benchmark_salt = MODULE_DEFAULT; module_ctx->module_benchmark_salt = MODULE_DEFAULT;
module_ctx->module_bridge_name = MODULE_DEFAULT;
module_ctx->module_bridge_type = MODULE_DEFAULT;
module_ctx->module_build_plain_postprocess = MODULE_DEFAULT; module_ctx->module_build_plain_postprocess = MODULE_DEFAULT;
module_ctx->module_deep_comp_kernel = MODULE_DEFAULT; module_ctx->module_deep_comp_kernel = MODULE_DEFAULT;
module_ctx->module_deprecated_notice = MODULE_DEFAULT; module_ctx->module_deprecated_notice = MODULE_DEFAULT;