X-Git-Url: https://www.flypig.org.uk/git/?a=blobdiff_plain;f=OpenCL%2Fm03710_a3.cl;h=b0d7c414c7313710f874bb2757bece76fc3252f2;hb=161a6eb4bc643d8e636e96eda613f5137d30da59;hp=91daea7865a1d57ed0020722941e8adb33218b1b;hpb=6893a9ad857a1d4fe63f1a92365802f5d9d3b39e;p=hashcat.git diff --git a/OpenCL/m03710_a3.cl b/OpenCL/m03710_a3.cl index 91daea7..b0d7c41 100644 --- a/OpenCL/m03710_a3.cl +++ b/OpenCL/m03710_a3.cl @@ -1,28 +1,35 @@ /** - * Author......: Jens Steube + * Authors.....: Jens Steube + * Gabriele Gristina + * magnum + * * License.....: MIT */ #define _MD5_ -#include "include/constants.h" -#include "include/kernel_vendor.h" - -#define DGST_R0 0 -#define DGST_R1 3 -#define DGST_R2 2 -#define DGST_R3 1 - -#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" - -#define uint_to_hex_lower8(i) l_bin2asc[(i)] - -static void m03710m (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], const u32 pw_len, __global pw_t *pws, __global gpu_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 bfs_cnt, const u32 digests_cnt, const u32 digests_offset, __local u32 l_bin2asc[256]) +#define NEW_SIMD_CODE + +#include "inc_vendor.cl" +#include "inc_hash_constants.h" +#include "inc_hash_functions.cl" +#include "inc_types.cl" +#include "inc_common.cl" +#include "inc_simd.cl" + +#if VECT_SIZE == 1 +#define uint_to_hex_lower8(i) (u32x) (l_bin2asc[(i)]) +#elif VECT_SIZE == 2 +#define uint_to_hex_lower8(i) (u32x) (l_bin2asc[(i).s0], l_bin2asc[(i).s1]) +#elif VECT_SIZE == 4 +#define uint_to_hex_lower8(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(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(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 + +void m03710m (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], const u32 pw_len, __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_scryptV0_buf, __global u32 *d_scryptV1_buf, __global u32 *d_scryptV2_buf, __global u32 *d_scryptV3_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, __local u32 *l_bin2asc) { /** * modifier @@ -36,32 +43,26 @@ static void m03710m (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], const u32 pw_le */ u32 salt_buf0[4]; + u32 salt_buf1[4]; + u32 salt_buf2[4]; + u32 salt_buf3[4]; salt_buf0[0] = salt_bufs[salt_pos].salt_buf[ 0]; salt_buf0[1] = salt_bufs[salt_pos].salt_buf[ 1]; salt_buf0[2] = salt_bufs[salt_pos].salt_buf[ 2]; salt_buf0[3] = salt_bufs[salt_pos].salt_buf[ 3]; - - u32 salt_buf1[4]; - salt_buf1[0] = salt_bufs[salt_pos].salt_buf[ 4]; salt_buf1[1] = salt_bufs[salt_pos].salt_buf[ 5]; salt_buf1[2] = salt_bufs[salt_pos].salt_buf[ 6]; salt_buf1[3] = salt_bufs[salt_pos].salt_buf[ 7]; - - u32 salt_buf2[4]; - - salt_buf2[0] = 0; - salt_buf2[1] = 0; - salt_buf2[2] = 0; - salt_buf2[3] = 0; - - u32 salt_buf3[4]; - - salt_buf3[0] = 0; - salt_buf3[1] = 0; - salt_buf3[2] = 0; - salt_buf3[3] = 0; + salt_buf2[0] = salt_bufs[salt_pos].salt_buf[ 8]; + salt_buf2[1] = salt_bufs[salt_pos].salt_buf[ 9]; + salt_buf2[2] = salt_bufs[salt_pos].salt_buf[10]; + salt_buf2[3] = salt_bufs[salt_pos].salt_buf[11]; + salt_buf3[0] = salt_bufs[salt_pos].salt_buf[12]; + salt_buf3[1] = salt_bufs[salt_pos].salt_buf[13]; + salt_buf3[2] = salt_bufs[salt_pos].salt_buf[14]; + salt_buf3[3] = salt_bufs[salt_pos].salt_buf[15]; const u32 salt_len = salt_bufs[salt_pos].salt_len; @@ -73,95 +74,116 @@ static void m03710m (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], const u32 pw_le u32 w0l = w0[0]; - for (u32 il_pos = 0; il_pos < bfs_cnt; il_pos++) + for (u32 il_pos = 0; il_pos < il_cnt; il_pos += VECT_SIZE) { - const u32 w0r = bfs_buf[il_pos].i; - - w0[0] = w0l | w0r; - - u32 a = MD5M_A; - u32 b = MD5M_B; - u32 c = MD5M_C; - u32 d = MD5M_D; - - MD5_STEP (MD5_Fo, a, b, c, d, w0[0], MD5C00, MD5S00); - MD5_STEP (MD5_Fo, d, a, b, c, w0[1], MD5C01, MD5S01); - MD5_STEP (MD5_Fo, c, d, a, b, w0[2], MD5C02, MD5S02); - MD5_STEP (MD5_Fo, b, c, d, a, w0[3], MD5C03, MD5S03); - MD5_STEP (MD5_Fo, a, b, c, d, w1[0], MD5C04, MD5S00); - MD5_STEP (MD5_Fo, d, a, b, c, w1[1], MD5C05, MD5S01); - MD5_STEP (MD5_Fo, c, d, a, b, w1[2], MD5C06, MD5S02); - MD5_STEP (MD5_Fo, b, c, d, a, w1[3], MD5C07, MD5S03); - MD5_STEP (MD5_Fo, a, b, c, d, w2[0], MD5C08, MD5S00); - MD5_STEP (MD5_Fo, d, a, b, c, w2[1], MD5C09, MD5S01); - MD5_STEP (MD5_Fo, c, d, a, b, w2[2], MD5C0a, MD5S02); - MD5_STEP (MD5_Fo, b, c, d, a, w2[3], MD5C0b, MD5S03); - MD5_STEP (MD5_Fo, a, b, c, d, w3[0], MD5C0c, MD5S00); - MD5_STEP (MD5_Fo, d, a, b, c, w3[1], MD5C0d, MD5S01); - MD5_STEP (MD5_Fo, c, d, a, b, w3[2], MD5C0e, MD5S02); - MD5_STEP (MD5_Fo, b, c, d, a, w3[3], MD5C0f, MD5S03); - - MD5_STEP (MD5_Go, a, b, c, d, w0[1], MD5C10, MD5S10); - MD5_STEP (MD5_Go, d, a, b, c, w1[2], MD5C11, MD5S11); - MD5_STEP (MD5_Go, c, d, a, b, w2[3], MD5C12, MD5S12); - MD5_STEP (MD5_Go, b, c, d, a, w0[0], MD5C13, MD5S13); - MD5_STEP (MD5_Go, a, b, c, d, w1[1], MD5C14, MD5S10); - MD5_STEP (MD5_Go, d, a, b, c, w2[2], MD5C15, MD5S11); - MD5_STEP (MD5_Go, c, d, a, b, w3[3], MD5C16, MD5S12); - MD5_STEP (MD5_Go, b, c, d, a, w1[0], MD5C17, MD5S13); - MD5_STEP (MD5_Go, a, b, c, d, w2[1], MD5C18, MD5S10); - MD5_STEP (MD5_Go, d, a, b, c, w3[2], MD5C19, MD5S11); - MD5_STEP (MD5_Go, c, d, a, b, w0[3], MD5C1a, MD5S12); - MD5_STEP (MD5_Go, b, c, d, a, w2[0], MD5C1b, MD5S13); - MD5_STEP (MD5_Go, a, b, c, d, w3[1], MD5C1c, MD5S10); - MD5_STEP (MD5_Go, d, a, b, c, w0[2], MD5C1d, MD5S11); - MD5_STEP (MD5_Go, c, d, a, b, w1[3], MD5C1e, MD5S12); - MD5_STEP (MD5_Go, b, c, d, a, w3[0], MD5C1f, MD5S13); - - MD5_STEP (MD5_H , a, b, c, d, w1[1], MD5C20, MD5S20); - MD5_STEP (MD5_H , d, a, b, c, w2[0], MD5C21, MD5S21); - MD5_STEP (MD5_H , c, d, a, b, w2[3], MD5C22, MD5S22); - MD5_STEP (MD5_H , b, c, d, a, w3[2], MD5C23, MD5S23); - MD5_STEP (MD5_H , a, b, c, d, w0[1], MD5C24, MD5S20); - MD5_STEP (MD5_H , d, a, b, c, w1[0], MD5C25, MD5S21); - MD5_STEP (MD5_H , c, d, a, b, w1[3], MD5C26, MD5S22); - MD5_STEP (MD5_H , b, c, d, a, w2[2], MD5C27, MD5S23); - MD5_STEP (MD5_H , a, b, c, d, w3[1], MD5C28, MD5S20); - MD5_STEP (MD5_H , d, a, b, c, w0[0], MD5C29, MD5S21); - MD5_STEP (MD5_H , c, d, a, b, w0[3], MD5C2a, MD5S22); - MD5_STEP (MD5_H , b, c, d, a, w1[2], MD5C2b, MD5S23); - MD5_STEP (MD5_H , a, b, c, d, w2[1], MD5C2c, MD5S20); - MD5_STEP (MD5_H , d, a, b, c, w3[0], MD5C2d, MD5S21); - MD5_STEP (MD5_H , c, d, a, b, w3[3], MD5C2e, MD5S22); - MD5_STEP (MD5_H , b, c, d, a, w0[2], MD5C2f, MD5S23); - - MD5_STEP (MD5_I , a, b, c, d, w0[0], MD5C30, MD5S30); - MD5_STEP (MD5_I , d, a, b, c, w1[3], MD5C31, MD5S31); - MD5_STEP (MD5_I , c, d, a, b, w3[2], MD5C32, MD5S32); - MD5_STEP (MD5_I , b, c, d, a, w1[1], MD5C33, MD5S33); - MD5_STEP (MD5_I , a, b, c, d, w3[0], MD5C34, MD5S30); - MD5_STEP (MD5_I , d, a, b, c, w0[3], MD5C35, MD5S31); - MD5_STEP (MD5_I , c, d, a, b, w2[2], MD5C36, MD5S32); - MD5_STEP (MD5_I , b, c, d, a, w0[1], MD5C37, MD5S33); - MD5_STEP (MD5_I , a, b, c, d, w2[0], MD5C38, MD5S30); - MD5_STEP (MD5_I , d, a, b, c, w3[3], MD5C39, MD5S31); - MD5_STEP (MD5_I , c, d, a, b, w1[2], MD5C3a, MD5S32); - MD5_STEP (MD5_I , b, c, d, a, w3[1], MD5C3b, MD5S33); - MD5_STEP (MD5_I , a, b, c, d, w1[0], MD5C3c, MD5S30); - MD5_STEP (MD5_I , d, a, b, c, w2[3], MD5C3d, MD5S31); - MD5_STEP (MD5_I , c, d, a, b, w0[2], MD5C3e, MD5S32); - MD5_STEP (MD5_I , b, c, d, a, w2[1], MD5C3f, MD5S33); + const u32x w0r = ix_create_bft (bfs_buf, il_pos); + + const u32x w0lr = w0l | w0r; + + u32x w0_t[4]; + u32x w1_t[4]; + u32x w2_t[4]; + u32x w3_t[4]; + + w0_t[0] = w0lr; + w0_t[1] = w0[1]; + w0_t[2] = w0[2]; + w0_t[3] = w0[3]; + w1_t[0] = w1[0]; + w1_t[1] = w1[1]; + w1_t[2] = w1[2]; + w1_t[3] = w1[3]; + w2_t[0] = w2[0]; + w2_t[1] = w2[1]; + w2_t[2] = w2[2]; + w2_t[3] = w2[3]; + w3_t[0] = w3[0]; + w3_t[1] = w3[1]; + w3_t[2] = w3[2]; + w3_t[3] = w3[3]; + + /** + * md5 + */ + + u32x a = MD5M_A; + u32x b = MD5M_B; + u32x c = MD5M_C; + u32x d = MD5M_D; + + MD5_STEP (MD5_Fo, a, b, c, d, w0_t[0], MD5C00, MD5S00); + MD5_STEP (MD5_Fo, d, a, b, c, w0_t[1], MD5C01, MD5S01); + MD5_STEP (MD5_Fo, c, d, a, b, w0_t[2], MD5C02, MD5S02); + MD5_STEP (MD5_Fo, b, c, d, a, w0_t[3], MD5C03, MD5S03); + MD5_STEP (MD5_Fo, a, b, c, d, w1_t[0], MD5C04, MD5S00); + MD5_STEP (MD5_Fo, d, a, b, c, w1_t[1], MD5C05, MD5S01); + MD5_STEP (MD5_Fo, c, d, a, b, w1_t[2], MD5C06, MD5S02); + MD5_STEP (MD5_Fo, b, c, d, a, w1_t[3], MD5C07, MD5S03); + MD5_STEP (MD5_Fo, a, b, c, d, w2_t[0], MD5C08, MD5S00); + MD5_STEP (MD5_Fo, d, a, b, c, w2_t[1], MD5C09, MD5S01); + MD5_STEP (MD5_Fo, c, d, a, b, w2_t[2], MD5C0a, MD5S02); + MD5_STEP (MD5_Fo, b, c, d, a, w2_t[3], MD5C0b, MD5S03); + MD5_STEP (MD5_Fo, a, b, c, d, w3_t[0], MD5C0c, MD5S00); + MD5_STEP (MD5_Fo, d, a, b, c, w3_t[1], MD5C0d, MD5S01); + MD5_STEP (MD5_Fo, c, d, a, b, w3_t[2], MD5C0e, MD5S02); + MD5_STEP (MD5_Fo, b, c, d, a, w3_t[3], MD5C0f, MD5S03); + + MD5_STEP (MD5_Go, a, b, c, d, w0_t[1], MD5C10, MD5S10); + MD5_STEP (MD5_Go, d, a, b, c, w1_t[2], MD5C11, MD5S11); + MD5_STEP (MD5_Go, c, d, a, b, w2_t[3], MD5C12, MD5S12); + MD5_STEP (MD5_Go, b, c, d, a, w0_t[0], MD5C13, MD5S13); + MD5_STEP (MD5_Go, a, b, c, d, w1_t[1], MD5C14, MD5S10); + MD5_STEP (MD5_Go, d, a, b, c, w2_t[2], MD5C15, MD5S11); + MD5_STEP (MD5_Go, c, d, a, b, w3_t[3], MD5C16, MD5S12); + MD5_STEP (MD5_Go, b, c, d, a, w1_t[0], MD5C17, MD5S13); + MD5_STEP (MD5_Go, a, b, c, d, w2_t[1], MD5C18, MD5S10); + MD5_STEP (MD5_Go, d, a, b, c, w3_t[2], MD5C19, MD5S11); + MD5_STEP (MD5_Go, c, d, a, b, w0_t[3], MD5C1a, MD5S12); + MD5_STEP (MD5_Go, b, c, d, a, w2_t[0], MD5C1b, MD5S13); + MD5_STEP (MD5_Go, a, b, c, d, w3_t[1], MD5C1c, MD5S10); + MD5_STEP (MD5_Go, d, a, b, c, w0_t[2], MD5C1d, MD5S11); + MD5_STEP (MD5_Go, c, d, a, b, w1_t[3], MD5C1e, MD5S12); + MD5_STEP (MD5_Go, b, c, d, a, w3_t[0], MD5C1f, MD5S13); + + MD5_STEP (MD5_H , a, b, c, d, w1_t[1], MD5C20, MD5S20); + MD5_STEP (MD5_H , d, a, b, c, w2_t[0], MD5C21, MD5S21); + MD5_STEP (MD5_H , c, d, a, b, w2_t[3], MD5C22, MD5S22); + MD5_STEP (MD5_H , b, c, d, a, w3_t[2], MD5C23, MD5S23); + MD5_STEP (MD5_H , a, b, c, d, w0_t[1], MD5C24, MD5S20); + MD5_STEP (MD5_H , d, a, b, c, w1_t[0], MD5C25, MD5S21); + MD5_STEP (MD5_H , c, d, a, b, w1_t[3], MD5C26, MD5S22); + MD5_STEP (MD5_H , b, c, d, a, w2_t[2], MD5C27, MD5S23); + MD5_STEP (MD5_H , a, b, c, d, w3_t[1], MD5C28, MD5S20); + MD5_STEP (MD5_H , d, a, b, c, w0_t[0], MD5C29, MD5S21); + MD5_STEP (MD5_H , c, d, a, b, w0_t[3], MD5C2a, MD5S22); + MD5_STEP (MD5_H , b, c, d, a, w1_t[2], MD5C2b, MD5S23); + MD5_STEP (MD5_H , a, b, c, d, w2_t[1], MD5C2c, MD5S20); + MD5_STEP (MD5_H , d, a, b, c, w3_t[0], MD5C2d, MD5S21); + MD5_STEP (MD5_H , c, d, a, b, w3_t[3], MD5C2e, MD5S22); + MD5_STEP (MD5_H , b, c, d, a, w0_t[2], MD5C2f, MD5S23); + + MD5_STEP (MD5_I , a, b, c, d, w0_t[0], MD5C30, MD5S30); + MD5_STEP (MD5_I , d, a, b, c, w1_t[3], MD5C31, MD5S31); + MD5_STEP (MD5_I , c, d, a, b, w3_t[2], MD5C32, MD5S32); + MD5_STEP (MD5_I , b, c, d, a, w1_t[1], MD5C33, MD5S33); + MD5_STEP (MD5_I , a, b, c, d, w3_t[0], MD5C34, MD5S30); + MD5_STEP (MD5_I , d, a, b, c, w0_t[3], MD5C35, MD5S31); + MD5_STEP (MD5_I , c, d, a, b, w2_t[2], MD5C36, MD5S32); + MD5_STEP (MD5_I , b, c, d, a, w0_t[1], MD5C37, MD5S33); + MD5_STEP (MD5_I , a, b, c, d, w2_t[0], MD5C38, MD5S30); + MD5_STEP (MD5_I , d, a, b, c, w3_t[3], MD5C39, MD5S31); + MD5_STEP (MD5_I , c, d, a, b, w1_t[2], MD5C3a, MD5S32); + MD5_STEP (MD5_I , b, c, d, a, w3_t[1], MD5C3b, MD5S33); + MD5_STEP (MD5_I , a, b, c, d, w1_t[0], MD5C3c, MD5S30); + MD5_STEP (MD5_I , d, a, b, c, w2_t[3], MD5C3d, MD5S31); + MD5_STEP (MD5_I , c, d, a, b, w0_t[2], MD5C3e, MD5S32); + MD5_STEP (MD5_I , b, c, d, a, w2_t[1], MD5C3f, MD5S33); a += MD5M_A; b += MD5M_B; c += MD5M_C; d += MD5M_D; - u32 w0_t[4]; - u32 w1_t[4]; - u32 w2_t[4]; - u32 w3_t[4]; - w0_t[0] = uint_to_hex_lower8 ((a >> 0) & 255) << 0 | uint_to_hex_lower8 ((a >> 8) & 255) << 16; w0_t[1] = uint_to_hex_lower8 ((a >> 16) & 255) << 0 @@ -178,12 +200,10 @@ static void m03710m (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], const u32 pw_le | uint_to_hex_lower8 ((d >> 8) & 255) << 16; w1_t[3] = uint_to_hex_lower8 ((d >> 16) & 255) << 0 | uint_to_hex_lower8 ((d >> 24) & 255) << 16; - - w2_t[0] = 0x00000080; + w2_t[0] = 0x80; w2_t[1] = 0; w2_t[2] = 0; w2_t[3] = 0; - w3_t[0] = 0; w3_t[1] = 0; w3_t[2] = 0; @@ -193,9 +213,10 @@ static void m03710m (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], const u32 pw_le * prepend salt */ - switch_buffer_by_offset (w0_t, w1_t, w2_t, w3_t, salt_len); + switch_buffer_by_offset_le (w0_t, w1_t, w2_t, w3_t, salt_len); w3_t[2] = pw_salt_len * 8; + w3_t[3] = 0; w0_t[0] |= salt_buf0[0]; w0_t[1] |= salt_buf0[1]; @@ -291,16 +312,11 @@ static void m03710m (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], const u32 pw_le MD5_STEP (MD5_I , c, d, a, b, w0_t[2], MD5C3e, MD5S32); MD5_STEP (MD5_I , b, c, d, a, w2_t[1], MD5C3f, MD5S33); - const u32 r0 = a; - const u32 r1 = d; - const u32 r2 = c; - const u32 r3 = b; - - #include COMPARE_M + COMPARE_M_SIMD (a, d, c, b); } } -static void m03710s (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], const u32 pw_len, __global pw_t *pws, __global gpu_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 bfs_cnt, const u32 digests_cnt, const u32 digests_offset, __local u32 l_bin2asc[256]) +void m03710s (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], const u32 pw_len, __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_scryptV0_buf, __global u32 *d_scryptV1_buf, __global u32 *d_scryptV2_buf, __global u32 *d_scryptV3_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, __local u32 *l_bin2asc) { /** * modifier @@ -309,149 +325,164 @@ static void m03710s (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], const u32 pw_le const u32 gid = get_global_id (0); const u32 lid = get_local_id (0); - /** - * 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] - }; - /** * salt */ u32 salt_buf0[4]; + u32 salt_buf1[4]; + u32 salt_buf2[4]; + u32 salt_buf3[4]; salt_buf0[0] = salt_bufs[salt_pos].salt_buf[ 0]; salt_buf0[1] = salt_bufs[salt_pos].salt_buf[ 1]; salt_buf0[2] = salt_bufs[salt_pos].salt_buf[ 2]; salt_buf0[3] = salt_bufs[salt_pos].salt_buf[ 3]; - - u32 salt_buf1[4]; - salt_buf1[0] = salt_bufs[salt_pos].salt_buf[ 4]; salt_buf1[1] = salt_bufs[salt_pos].salt_buf[ 5]; salt_buf1[2] = salt_bufs[salt_pos].salt_buf[ 6]; salt_buf1[3] = salt_bufs[salt_pos].salt_buf[ 7]; - - u32 salt_buf2[4]; - - salt_buf2[0] = 0; - salt_buf2[1] = 0; - salt_buf2[2] = 0; - salt_buf2[3] = 0; - - u32 salt_buf3[4]; - - salt_buf3[0] = 0; - salt_buf3[1] = 0; - salt_buf3[2] = 0; - salt_buf3[3] = 0; + salt_buf2[0] = salt_bufs[salt_pos].salt_buf[ 8]; + salt_buf2[1] = salt_bufs[salt_pos].salt_buf[ 9]; + salt_buf2[2] = salt_bufs[salt_pos].salt_buf[10]; + salt_buf2[3] = salt_bufs[salt_pos].salt_buf[11]; + salt_buf3[0] = salt_bufs[salt_pos].salt_buf[12]; + salt_buf3[1] = salt_bufs[salt_pos].salt_buf[13]; + salt_buf3[2] = salt_bufs[salt_pos].salt_buf[14]; + salt_buf3[3] = salt_bufs[salt_pos].salt_buf[15]; const u32 salt_len = salt_bufs[salt_pos].salt_len; const u32 pw_salt_len = 32 + salt_len; + /** + * 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] + }; + /** * loop */ u32 w0l = w0[0]; - for (u32 il_pos = 0; il_pos < bfs_cnt; il_pos++) + for (u32 il_pos = 0; il_pos < il_cnt; il_pos += VECT_SIZE) { - const u32 w0r = bfs_buf[il_pos].i; - - w0[0] = w0l | w0r; - - u32 a = MD5M_A; - u32 b = MD5M_B; - u32 c = MD5M_C; - u32 d = MD5M_D; - - MD5_STEP (MD5_Fo, a, b, c, d, w0[0], MD5C00, MD5S00); - MD5_STEP (MD5_Fo, d, a, b, c, w0[1], MD5C01, MD5S01); - MD5_STEP (MD5_Fo, c, d, a, b, w0[2], MD5C02, MD5S02); - MD5_STEP (MD5_Fo, b, c, d, a, w0[3], MD5C03, MD5S03); - MD5_STEP (MD5_Fo, a, b, c, d, w1[0], MD5C04, MD5S00); - MD5_STEP (MD5_Fo, d, a, b, c, w1[1], MD5C05, MD5S01); - MD5_STEP (MD5_Fo, c, d, a, b, w1[2], MD5C06, MD5S02); - MD5_STEP (MD5_Fo, b, c, d, a, w1[3], MD5C07, MD5S03); - MD5_STEP (MD5_Fo, a, b, c, d, w2[0], MD5C08, MD5S00); - MD5_STEP (MD5_Fo, d, a, b, c, w2[1], MD5C09, MD5S01); - MD5_STEP (MD5_Fo, c, d, a, b, w2[2], MD5C0a, MD5S02); - MD5_STEP (MD5_Fo, b, c, d, a, w2[3], MD5C0b, MD5S03); - MD5_STEP (MD5_Fo, a, b, c, d, w3[0], MD5C0c, MD5S00); - MD5_STEP (MD5_Fo, d, a, b, c, w3[1], MD5C0d, MD5S01); - MD5_STEP (MD5_Fo, c, d, a, b, w3[2], MD5C0e, MD5S02); - MD5_STEP (MD5_Fo, b, c, d, a, w3[3], MD5C0f, MD5S03); - - MD5_STEP (MD5_Go, a, b, c, d, w0[1], MD5C10, MD5S10); - MD5_STEP (MD5_Go, d, a, b, c, w1[2], MD5C11, MD5S11); - MD5_STEP (MD5_Go, c, d, a, b, w2[3], MD5C12, MD5S12); - MD5_STEP (MD5_Go, b, c, d, a, w0[0], MD5C13, MD5S13); - MD5_STEP (MD5_Go, a, b, c, d, w1[1], MD5C14, MD5S10); - MD5_STEP (MD5_Go, d, a, b, c, w2[2], MD5C15, MD5S11); - MD5_STEP (MD5_Go, c, d, a, b, w3[3], MD5C16, MD5S12); - MD5_STEP (MD5_Go, b, c, d, a, w1[0], MD5C17, MD5S13); - MD5_STEP (MD5_Go, a, b, c, d, w2[1], MD5C18, MD5S10); - MD5_STEP (MD5_Go, d, a, b, c, w3[2], MD5C19, MD5S11); - MD5_STEP (MD5_Go, c, d, a, b, w0[3], MD5C1a, MD5S12); - MD5_STEP (MD5_Go, b, c, d, a, w2[0], MD5C1b, MD5S13); - MD5_STEP (MD5_Go, a, b, c, d, w3[1], MD5C1c, MD5S10); - MD5_STEP (MD5_Go, d, a, b, c, w0[2], MD5C1d, MD5S11); - MD5_STEP (MD5_Go, c, d, a, b, w1[3], MD5C1e, MD5S12); - MD5_STEP (MD5_Go, b, c, d, a, w3[0], MD5C1f, MD5S13); - - MD5_STEP (MD5_H , a, b, c, d, w1[1], MD5C20, MD5S20); - MD5_STEP (MD5_H , d, a, b, c, w2[0], MD5C21, MD5S21); - MD5_STEP (MD5_H , c, d, a, b, w2[3], MD5C22, MD5S22); - MD5_STEP (MD5_H , b, c, d, a, w3[2], MD5C23, MD5S23); - MD5_STEP (MD5_H , a, b, c, d, w0[1], MD5C24, MD5S20); - MD5_STEP (MD5_H , d, a, b, c, w1[0], MD5C25, MD5S21); - MD5_STEP (MD5_H , c, d, a, b, w1[3], MD5C26, MD5S22); - MD5_STEP (MD5_H , b, c, d, a, w2[2], MD5C27, MD5S23); - MD5_STEP (MD5_H , a, b, c, d, w3[1], MD5C28, MD5S20); - MD5_STEP (MD5_H , d, a, b, c, w0[0], MD5C29, MD5S21); - MD5_STEP (MD5_H , c, d, a, b, w0[3], MD5C2a, MD5S22); - MD5_STEP (MD5_H , b, c, d, a, w1[2], MD5C2b, MD5S23); - MD5_STEP (MD5_H , a, b, c, d, w2[1], MD5C2c, MD5S20); - MD5_STEP (MD5_H , d, a, b, c, w3[0], MD5C2d, MD5S21); - MD5_STEP (MD5_H , c, d, a, b, w3[3], MD5C2e, MD5S22); - MD5_STEP (MD5_H , b, c, d, a, w0[2], MD5C2f, MD5S23); - - MD5_STEP (MD5_I , a, b, c, d, w0[0], MD5C30, MD5S30); - MD5_STEP (MD5_I , d, a, b, c, w1[3], MD5C31, MD5S31); - MD5_STEP (MD5_I , c, d, a, b, w3[2], MD5C32, MD5S32); - MD5_STEP (MD5_I , b, c, d, a, w1[1], MD5C33, MD5S33); - MD5_STEP (MD5_I , a, b, c, d, w3[0], MD5C34, MD5S30); - MD5_STEP (MD5_I , d, a, b, c, w0[3], MD5C35, MD5S31); - MD5_STEP (MD5_I , c, d, a, b, w2[2], MD5C36, MD5S32); - MD5_STEP (MD5_I , b, c, d, a, w0[1], MD5C37, MD5S33); - MD5_STEP (MD5_I , a, b, c, d, w2[0], MD5C38, MD5S30); - MD5_STEP (MD5_I , d, a, b, c, w3[3], MD5C39, MD5S31); - MD5_STEP (MD5_I , c, d, a, b, w1[2], MD5C3a, MD5S32); - MD5_STEP (MD5_I , b, c, d, a, w3[1], MD5C3b, MD5S33); - MD5_STEP (MD5_I , a, b, c, d, w1[0], MD5C3c, MD5S30); - MD5_STEP (MD5_I , d, a, b, c, w2[3], MD5C3d, MD5S31); - MD5_STEP (MD5_I , c, d, a, b, w0[2], MD5C3e, MD5S32); - MD5_STEP (MD5_I , b, c, d, a, w2[1], MD5C3f, MD5S33); + const u32x w0r = ix_create_bft (bfs_buf, il_pos); + + const u32x w0lr = w0l | w0r; + + u32x w0_t[4]; + u32x w1_t[4]; + u32x w2_t[4]; + u32x w3_t[4]; + + w0_t[0] = w0lr; + w0_t[1] = w0[1]; + w0_t[2] = w0[2]; + w0_t[3] = w0[3]; + w1_t[0] = w1[0]; + w1_t[1] = w1[1]; + w1_t[2] = w1[2]; + w1_t[3] = w1[3]; + w2_t[0] = w2[0]; + w2_t[1] = w2[1]; + w2_t[2] = w2[2]; + w2_t[3] = w2[3]; + w3_t[0] = w3[0]; + w3_t[1] = w3[1]; + w3_t[2] = w3[2]; + w3_t[3] = w3[3]; + + /** + * md5 + */ + + u32x a = MD5M_A; + u32x b = MD5M_B; + u32x c = MD5M_C; + u32x d = MD5M_D; + + MD5_STEP (MD5_Fo, a, b, c, d, w0_t[0], MD5C00, MD5S00); + MD5_STEP (MD5_Fo, d, a, b, c, w0_t[1], MD5C01, MD5S01); + MD5_STEP (MD5_Fo, c, d, a, b, w0_t[2], MD5C02, MD5S02); + MD5_STEP (MD5_Fo, b, c, d, a, w0_t[3], MD5C03, MD5S03); + MD5_STEP (MD5_Fo, a, b, c, d, w1_t[0], MD5C04, MD5S00); + MD5_STEP (MD5_Fo, d, a, b, c, w1_t[1], MD5C05, MD5S01); + MD5_STEP (MD5_Fo, c, d, a, b, w1_t[2], MD5C06, MD5S02); + MD5_STEP (MD5_Fo, b, c, d, a, w1_t[3], MD5C07, MD5S03); + MD5_STEP (MD5_Fo, a, b, c, d, w2_t[0], MD5C08, MD5S00); + MD5_STEP (MD5_Fo, d, a, b, c, w2_t[1], MD5C09, MD5S01); + MD5_STEP (MD5_Fo, c, d, a, b, w2_t[2], MD5C0a, MD5S02); + MD5_STEP (MD5_Fo, b, c, d, a, w2_t[3], MD5C0b, MD5S03); + MD5_STEP (MD5_Fo, a, b, c, d, w3_t[0], MD5C0c, MD5S00); + MD5_STEP (MD5_Fo, d, a, b, c, w3_t[1], MD5C0d, MD5S01); + MD5_STEP (MD5_Fo, c, d, a, b, w3_t[2], MD5C0e, MD5S02); + MD5_STEP (MD5_Fo, b, c, d, a, w3_t[3], MD5C0f, MD5S03); + + MD5_STEP (MD5_Go, a, b, c, d, w0_t[1], MD5C10, MD5S10); + MD5_STEP (MD5_Go, d, a, b, c, w1_t[2], MD5C11, MD5S11); + MD5_STEP (MD5_Go, c, d, a, b, w2_t[3], MD5C12, MD5S12); + MD5_STEP (MD5_Go, b, c, d, a, w0_t[0], MD5C13, MD5S13); + MD5_STEP (MD5_Go, a, b, c, d, w1_t[1], MD5C14, MD5S10); + MD5_STEP (MD5_Go, d, a, b, c, w2_t[2], MD5C15, MD5S11); + MD5_STEP (MD5_Go, c, d, a, b, w3_t[3], MD5C16, MD5S12); + MD5_STEP (MD5_Go, b, c, d, a, w1_t[0], MD5C17, MD5S13); + MD5_STEP (MD5_Go, a, b, c, d, w2_t[1], MD5C18, MD5S10); + MD5_STEP (MD5_Go, d, a, b, c, w3_t[2], MD5C19, MD5S11); + MD5_STEP (MD5_Go, c, d, a, b, w0_t[3], MD5C1a, MD5S12); + MD5_STEP (MD5_Go, b, c, d, a, w2_t[0], MD5C1b, MD5S13); + MD5_STEP (MD5_Go, a, b, c, d, w3_t[1], MD5C1c, MD5S10); + MD5_STEP (MD5_Go, d, a, b, c, w0_t[2], MD5C1d, MD5S11); + MD5_STEP (MD5_Go, c, d, a, b, w1_t[3], MD5C1e, MD5S12); + MD5_STEP (MD5_Go, b, c, d, a, w3_t[0], MD5C1f, MD5S13); + + MD5_STEP (MD5_H , a, b, c, d, w1_t[1], MD5C20, MD5S20); + MD5_STEP (MD5_H , d, a, b, c, w2_t[0], MD5C21, MD5S21); + MD5_STEP (MD5_H , c, d, a, b, w2_t[3], MD5C22, MD5S22); + MD5_STEP (MD5_H , b, c, d, a, w3_t[2], MD5C23, MD5S23); + MD5_STEP (MD5_H , a, b, c, d, w0_t[1], MD5C24, MD5S20); + MD5_STEP (MD5_H , d, a, b, c, w1_t[0], MD5C25, MD5S21); + MD5_STEP (MD5_H , c, d, a, b, w1_t[3], MD5C26, MD5S22); + MD5_STEP (MD5_H , b, c, d, a, w2_t[2], MD5C27, MD5S23); + MD5_STEP (MD5_H , a, b, c, d, w3_t[1], MD5C28, MD5S20); + MD5_STEP (MD5_H , d, a, b, c, w0_t[0], MD5C29, MD5S21); + MD5_STEP (MD5_H , c, d, a, b, w0_t[3], MD5C2a, MD5S22); + MD5_STEP (MD5_H , b, c, d, a, w1_t[2], MD5C2b, MD5S23); + MD5_STEP (MD5_H , a, b, c, d, w2_t[1], MD5C2c, MD5S20); + MD5_STEP (MD5_H , d, a, b, c, w3_t[0], MD5C2d, MD5S21); + MD5_STEP (MD5_H , c, d, a, b, w3_t[3], MD5C2e, MD5S22); + MD5_STEP (MD5_H , b, c, d, a, w0_t[2], MD5C2f, MD5S23); + + MD5_STEP (MD5_I , a, b, c, d, w0_t[0], MD5C30, MD5S30); + MD5_STEP (MD5_I , d, a, b, c, w1_t[3], MD5C31, MD5S31); + MD5_STEP (MD5_I , c, d, a, b, w3_t[2], MD5C32, MD5S32); + MD5_STEP (MD5_I , b, c, d, a, w1_t[1], MD5C33, MD5S33); + MD5_STEP (MD5_I , a, b, c, d, w3_t[0], MD5C34, MD5S30); + MD5_STEP (MD5_I , d, a, b, c, w0_t[3], MD5C35, MD5S31); + MD5_STEP (MD5_I , c, d, a, b, w2_t[2], MD5C36, MD5S32); + MD5_STEP (MD5_I , b, c, d, a, w0_t[1], MD5C37, MD5S33); + MD5_STEP (MD5_I , a, b, c, d, w2_t[0], MD5C38, MD5S30); + MD5_STEP (MD5_I , d, a, b, c, w3_t[3], MD5C39, MD5S31); + MD5_STEP (MD5_I , c, d, a, b, w1_t[2], MD5C3a, MD5S32); + MD5_STEP (MD5_I , b, c, d, a, w3_t[1], MD5C3b, MD5S33); + MD5_STEP (MD5_I , a, b, c, d, w1_t[0], MD5C3c, MD5S30); + MD5_STEP (MD5_I , d, a, b, c, w2_t[3], MD5C3d, MD5S31); + MD5_STEP (MD5_I , c, d, a, b, w0_t[2], MD5C3e, MD5S32); + MD5_STEP (MD5_I , b, c, d, a, w2_t[1], MD5C3f, MD5S33); a += MD5M_A; b += MD5M_B; c += MD5M_C; d += MD5M_D; - u32 w0_t[4]; - u32 w1_t[4]; - u32 w2_t[4]; - u32 w3_t[4]; - w0_t[0] = uint_to_hex_lower8 ((a >> 0) & 255) << 0 | uint_to_hex_lower8 ((a >> 8) & 255) << 16; w0_t[1] = uint_to_hex_lower8 ((a >> 16) & 255) << 0 @@ -468,12 +499,10 @@ static void m03710s (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], const u32 pw_le | uint_to_hex_lower8 ((d >> 8) & 255) << 16; w1_t[3] = uint_to_hex_lower8 ((d >> 16) & 255) << 0 | uint_to_hex_lower8 ((d >> 24) & 255) << 16; - - w2_t[0] = 0x00000080; + w2_t[0] = 0x80; w2_t[1] = 0; w2_t[2] = 0; w2_t[3] = 0; - w3_t[0] = 0; w3_t[1] = 0; w3_t[2] = 0; @@ -483,9 +512,10 @@ static void m03710s (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], const u32 pw_le * prepend salt */ - switch_buffer_by_offset (w0_t, w1_t, w2_t, w3_t, salt_len); + switch_buffer_by_offset_le (w0_t, w1_t, w2_t, w3_t, salt_len); w3_t[2] = pw_salt_len * 8; + w3_t[3] = 0; w0_t[0] |= salt_buf0[0]; w0_t[1] |= salt_buf0[1]; @@ -577,32 +607,49 @@ static void m03710s (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], const u32 pw_le MD5_STEP (MD5_I , c, d, a, b, w1_t[2], MD5C3a, MD5S32); MD5_STEP (MD5_I , b, c, d, a, w3_t[1], MD5C3b, MD5S33); MD5_STEP (MD5_I , a, b, c, d, w1_t[0], MD5C3c, MD5S30); + + if (MATCHES_NONE_VS (a, search[0])) continue; + MD5_STEP (MD5_I , d, a, b, c, w2_t[3], MD5C3d, MD5S31); MD5_STEP (MD5_I , c, d, a, b, w0_t[2], MD5C3e, MD5S32); MD5_STEP (MD5_I , b, c, d, a, w2_t[1], MD5C3f, MD5S33); - const u32 r0 = a; - const u32 r1 = d; - const u32 r2 = c; - const u32 r3 = b; - - #include COMPARE_S + COMPARE_S_SIMD (a, d, c, b); } } -__kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m03710_m04 (__global pw_t *pws, __global gpu_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 bfs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max) +__kernel void m03710_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_scryptV0_buf, __global u32 *d_scryptV1_buf, __global u32 *d_scryptV2_buf, __global u32 *d_scryptV3_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 */ const u32 gid = get_global_id (0); + const u32 lid = get_local_id (0); + const u32 lsz = get_local_size (0); /** - * modifier + * bin2asc table */ - const u32 lid = get_local_id (0); + __local u32 l_bin2asc[256]; + + for (u32 i = lid; i < 256; i += lsz) + { + const u32 i0 = (i >> 0) & 15; + const u32 i1 = (i >> 4) & 15; + + l_bin2asc[i] = ((i0 < 10) ? '0' + i0 : 'a' - 10 + i0) << 8 + | ((i1 < 10) ? '0' + i1 : 'a' - 10 + i1) << 0; + } + + barrier (CLK_LOCAL_MEM_FENCE); + + if (gid >= gid_max) return; + + /** + * modifier + */ u32 w0[4]; @@ -634,62 +681,27 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m03710_m04 (__glo const u32 pw_len = pws[gid].pw_len; - /** - * bin2asc table - */ - - __local u32 l_bin2asc[256]; - - const u32 lid4 = lid * 4; - - const u32 lid40 = lid4 + 0; - const u32 lid41 = lid4 + 1; - const u32 lid42 = lid4 + 2; - const u32 lid43 = lid4 + 3; - - const u32 v400 = (lid40 >> 0) & 15; - const u32 v401 = (lid40 >> 4) & 15; - const u32 v410 = (lid41 >> 0) & 15; - const u32 v411 = (lid41 >> 4) & 15; - const u32 v420 = (lid42 >> 0) & 15; - const u32 v421 = (lid42 >> 4) & 15; - const u32 v430 = (lid43 >> 0) & 15; - const u32 v431 = (lid43 >> 4) & 15; - - l_bin2asc[lid40] = ((v400 < 10) ? '0' + v400 : 'a' - 10 + v400) << 8 - | ((v401 < 10) ? '0' + v401 : 'a' - 10 + v401) << 0; - l_bin2asc[lid41] = ((v410 < 10) ? '0' + v410 : 'a' - 10 + v410) << 8 - | ((v411 < 10) ? '0' + v411 : 'a' - 10 + v411) << 0; - l_bin2asc[lid42] = ((v420 < 10) ? '0' + v420 : 'a' - 10 + v420) << 8 - | ((v421 < 10) ? '0' + v421 : 'a' - 10 + v421) << 0; - l_bin2asc[lid43] = ((v430 < 10) ? '0' + v430 : 'a' - 10 + v430) << 8 - | ((v431 < 10) ? '0' + v431 : 'a' - 10 + v431) << 0; - - barrier (CLK_LOCAL_MEM_FENCE); - - if (gid >= gid_max) return; - /** * main */ - m03710m (w0, w1, w2, w3, pw_len, pws, rules_buf, combs_buf, bfs_buf, tmps, hooks, bitmaps_buf_s1_a, bitmaps_buf_s1_b, bitmaps_buf_s1_c, bitmaps_buf_s1_d, bitmaps_buf_s2_a, bitmaps_buf_s2_b, bitmaps_buf_s2_c, bitmaps_buf_s2_d, plains_buf, digests_buf, hashes_shown, salt_bufs, esalt_bufs, d_return_buf, d_scryptV_buf, bitmap_mask, bitmap_shift1, bitmap_shift2, salt_pos, loop_pos, loop_cnt, bfs_cnt, digests_cnt, digests_offset, l_bin2asc); + m03710m (w0, w1, w2, w3, pw_len, pws, rules_buf, combs_buf, bfs_buf, tmps, hooks, bitmaps_buf_s1_a, bitmaps_buf_s1_b, bitmaps_buf_s1_c, bitmaps_buf_s1_d, bitmaps_buf_s2_a, bitmaps_buf_s2_b, bitmaps_buf_s2_c, bitmaps_buf_s2_d, plains_buf, digests_buf, hashes_shown, salt_bufs, esalt_bufs, d_return_buf, d_scryptV0_buf, d_scryptV1_buf, d_scryptV2_buf, d_scryptV3_buf, bitmap_mask, bitmap_shift1, bitmap_shift2, salt_pos, loop_pos, loop_cnt, il_cnt, digests_cnt, digests_offset, l_bin2asc); } -__kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m03710_m08 (__global pw_t *pws, __global gpu_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 bfs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max) +__kernel void m03710_m08 (__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_scryptV0_buf, __global u32 *d_scryptV1_buf, __global u32 *d_scryptV2_buf, __global u32 *d_scryptV3_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 */ const u32 gid = get_global_id (0); + const u32 lid = get_local_id (0); + const u32 lsz = get_local_size (0); /** * modifier */ - const u32 lid = get_local_id (0); - u32 w0[4]; w0[0] = pws[gid].i[ 0]; @@ -726,30 +738,14 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m03710_m08 (__glo __local u32 l_bin2asc[256]; - const u32 lid4 = lid * 4; - - const u32 lid40 = lid4 + 0; - const u32 lid41 = lid4 + 1; - const u32 lid42 = lid4 + 2; - const u32 lid43 = lid4 + 3; - - const u32 v400 = (lid40 >> 0) & 15; - const u32 v401 = (lid40 >> 4) & 15; - const u32 v410 = (lid41 >> 0) & 15; - const u32 v411 = (lid41 >> 4) & 15; - const u32 v420 = (lid42 >> 0) & 15; - const u32 v421 = (lid42 >> 4) & 15; - const u32 v430 = (lid43 >> 0) & 15; - const u32 v431 = (lid43 >> 4) & 15; - - l_bin2asc[lid40] = ((v400 < 10) ? '0' + v400 : 'a' - 10 + v400) << 8 - | ((v401 < 10) ? '0' + v401 : 'a' - 10 + v401) << 0; - l_bin2asc[lid41] = ((v410 < 10) ? '0' + v410 : 'a' - 10 + v410) << 8 - | ((v411 < 10) ? '0' + v411 : 'a' - 10 + v411) << 0; - l_bin2asc[lid42] = ((v420 < 10) ? '0' + v420 : 'a' - 10 + v420) << 8 - | ((v421 < 10) ? '0' + v421 : 'a' - 10 + v421) << 0; - l_bin2asc[lid43] = ((v430 < 10) ? '0' + v430 : 'a' - 10 + v430) << 8 - | ((v431 < 10) ? '0' + v431 : 'a' - 10 + v431) << 0; + for (u32 i = lid; i < 256; i += lsz) + { + const u32 i0 = (i >> 0) & 15; + const u32 i1 = (i >> 4) & 15; + + l_bin2asc[i] = ((i0 < 10) ? '0' + i0 : 'a' - 10 + i0) << 8 + | ((i1 < 10) ? '0' + i1 : 'a' - 10 + i1) << 0; + } barrier (CLK_LOCAL_MEM_FENCE); @@ -759,23 +755,23 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m03710_m08 (__glo * main */ - m03710m (w0, w1, w2, w3, pw_len, pws, rules_buf, combs_buf, bfs_buf, tmps, hooks, bitmaps_buf_s1_a, bitmaps_buf_s1_b, bitmaps_buf_s1_c, bitmaps_buf_s1_d, bitmaps_buf_s2_a, bitmaps_buf_s2_b, bitmaps_buf_s2_c, bitmaps_buf_s2_d, plains_buf, digests_buf, hashes_shown, salt_bufs, esalt_bufs, d_return_buf, d_scryptV_buf, bitmap_mask, bitmap_shift1, bitmap_shift2, salt_pos, loop_pos, loop_cnt, bfs_cnt, digests_cnt, digests_offset, l_bin2asc); + m03710m (w0, w1, w2, w3, pw_len, pws, rules_buf, combs_buf, bfs_buf, tmps, hooks, bitmaps_buf_s1_a, bitmaps_buf_s1_b, bitmaps_buf_s1_c, bitmaps_buf_s1_d, bitmaps_buf_s2_a, bitmaps_buf_s2_b, bitmaps_buf_s2_c, bitmaps_buf_s2_d, plains_buf, digests_buf, hashes_shown, salt_bufs, esalt_bufs, d_return_buf, d_scryptV0_buf, d_scryptV1_buf, d_scryptV2_buf, d_scryptV3_buf, bitmap_mask, bitmap_shift1, bitmap_shift2, salt_pos, loop_pos, loop_cnt, il_cnt, digests_cnt, digests_offset, l_bin2asc); } -__kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m03710_m16 (__global pw_t *pws, __global gpu_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 bfs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max) +__kernel void m03710_m16 (__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_scryptV0_buf, __global u32 *d_scryptV1_buf, __global u32 *d_scryptV2_buf, __global u32 *d_scryptV3_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 */ const u32 gid = get_global_id (0); + const u32 lid = get_local_id (0); + const u32 lsz = get_local_size (0); /** * modifier */ - const u32 lid = get_local_id (0); - u32 w0[4]; w0[0] = pws[gid].i[ 0]; @@ -812,30 +808,14 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m03710_m16 (__glo __local u32 l_bin2asc[256]; - const u32 lid4 = lid * 4; - - const u32 lid40 = lid4 + 0; - const u32 lid41 = lid4 + 1; - const u32 lid42 = lid4 + 2; - const u32 lid43 = lid4 + 3; - - const u32 v400 = (lid40 >> 0) & 15; - const u32 v401 = (lid40 >> 4) & 15; - const u32 v410 = (lid41 >> 0) & 15; - const u32 v411 = (lid41 >> 4) & 15; - const u32 v420 = (lid42 >> 0) & 15; - const u32 v421 = (lid42 >> 4) & 15; - const u32 v430 = (lid43 >> 0) & 15; - const u32 v431 = (lid43 >> 4) & 15; - - l_bin2asc[lid40] = ((v400 < 10) ? '0' + v400 : 'a' - 10 + v400) << 8 - | ((v401 < 10) ? '0' + v401 : 'a' - 10 + v401) << 0; - l_bin2asc[lid41] = ((v410 < 10) ? '0' + v410 : 'a' - 10 + v410) << 8 - | ((v411 < 10) ? '0' + v411 : 'a' - 10 + v411) << 0; - l_bin2asc[lid42] = ((v420 < 10) ? '0' + v420 : 'a' - 10 + v420) << 8 - | ((v421 < 10) ? '0' + v421 : 'a' - 10 + v421) << 0; - l_bin2asc[lid43] = ((v430 < 10) ? '0' + v430 : 'a' - 10 + v430) << 8 - | ((v431 < 10) ? '0' + v431 : 'a' - 10 + v431) << 0; + for (u32 i = lid; i < 256; i += lsz) + { + const u32 i0 = (i >> 0) & 15; + const u32 i1 = (i >> 4) & 15; + + l_bin2asc[i] = ((i0 < 10) ? '0' + i0 : 'a' - 10 + i0) << 8 + | ((i1 < 10) ? '0' + i1 : 'a' - 10 + i1) << 0; + } barrier (CLK_LOCAL_MEM_FENCE); @@ -845,23 +825,23 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m03710_m16 (__glo * main */ - m03710m (w0, w1, w2, w3, pw_len, pws, rules_buf, combs_buf, bfs_buf, tmps, hooks, bitmaps_buf_s1_a, bitmaps_buf_s1_b, bitmaps_buf_s1_c, bitmaps_buf_s1_d, bitmaps_buf_s2_a, bitmaps_buf_s2_b, bitmaps_buf_s2_c, bitmaps_buf_s2_d, plains_buf, digests_buf, hashes_shown, salt_bufs, esalt_bufs, d_return_buf, d_scryptV_buf, bitmap_mask, bitmap_shift1, bitmap_shift2, salt_pos, loop_pos, loop_cnt, bfs_cnt, digests_cnt, digests_offset, l_bin2asc); + m03710m (w0, w1, w2, w3, pw_len, pws, rules_buf, combs_buf, bfs_buf, tmps, hooks, bitmaps_buf_s1_a, bitmaps_buf_s1_b, bitmaps_buf_s1_c, bitmaps_buf_s1_d, bitmaps_buf_s2_a, bitmaps_buf_s2_b, bitmaps_buf_s2_c, bitmaps_buf_s2_d, plains_buf, digests_buf, hashes_shown, salt_bufs, esalt_bufs, d_return_buf, d_scryptV0_buf, d_scryptV1_buf, d_scryptV2_buf, d_scryptV3_buf, bitmap_mask, bitmap_shift1, bitmap_shift2, salt_pos, loop_pos, loop_cnt, il_cnt, digests_cnt, digests_offset, l_bin2asc); } -__kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m03710_s04 (__global pw_t *pws, __global gpu_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 bfs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max) +__kernel void m03710_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_scryptV0_buf, __global u32 *d_scryptV1_buf, __global u32 *d_scryptV2_buf, __global u32 *d_scryptV3_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 */ const u32 gid = get_global_id (0); + const u32 lid = get_local_id (0); + const u32 lsz = get_local_size (0); /** * modifier */ - const u32 lid = get_local_id (0); - u32 w0[4]; w0[0] = pws[gid].i[ 0]; @@ -898,30 +878,14 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m03710_s04 (__glo __local u32 l_bin2asc[256]; - const u32 lid4 = lid * 4; - - const u32 lid40 = lid4 + 0; - const u32 lid41 = lid4 + 1; - const u32 lid42 = lid4 + 2; - const u32 lid43 = lid4 + 3; - - const u32 v400 = (lid40 >> 0) & 15; - const u32 v401 = (lid40 >> 4) & 15; - const u32 v410 = (lid41 >> 0) & 15; - const u32 v411 = (lid41 >> 4) & 15; - const u32 v420 = (lid42 >> 0) & 15; - const u32 v421 = (lid42 >> 4) & 15; - const u32 v430 = (lid43 >> 0) & 15; - const u32 v431 = (lid43 >> 4) & 15; - - l_bin2asc[lid40] = ((v400 < 10) ? '0' + v400 : 'a' - 10 + v400) << 8 - | ((v401 < 10) ? '0' + v401 : 'a' - 10 + v401) << 0; - l_bin2asc[lid41] = ((v410 < 10) ? '0' + v410 : 'a' - 10 + v410) << 8 - | ((v411 < 10) ? '0' + v411 : 'a' - 10 + v411) << 0; - l_bin2asc[lid42] = ((v420 < 10) ? '0' + v420 : 'a' - 10 + v420) << 8 - | ((v421 < 10) ? '0' + v421 : 'a' - 10 + v421) << 0; - l_bin2asc[lid43] = ((v430 < 10) ? '0' + v430 : 'a' - 10 + v430) << 8 - | ((v431 < 10) ? '0' + v431 : 'a' - 10 + v431) << 0; + for (u32 i = lid; i < 256; i += lsz) + { + const u32 i0 = (i >> 0) & 15; + const u32 i1 = (i >> 4) & 15; + + l_bin2asc[i] = ((i0 < 10) ? '0' + i0 : 'a' - 10 + i0) << 8 + | ((i1 < 10) ? '0' + i1 : 'a' - 10 + i1) << 0; + } barrier (CLK_LOCAL_MEM_FENCE); @@ -931,23 +895,23 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m03710_s04 (__glo * main */ - m03710s (w0, w1, w2, w3, pw_len, pws, rules_buf, combs_buf, bfs_buf, tmps, hooks, bitmaps_buf_s1_a, bitmaps_buf_s1_b, bitmaps_buf_s1_c, bitmaps_buf_s1_d, bitmaps_buf_s2_a, bitmaps_buf_s2_b, bitmaps_buf_s2_c, bitmaps_buf_s2_d, plains_buf, digests_buf, hashes_shown, salt_bufs, esalt_bufs, d_return_buf, d_scryptV_buf, bitmap_mask, bitmap_shift1, bitmap_shift2, salt_pos, loop_pos, loop_cnt, bfs_cnt, digests_cnt, digests_offset, l_bin2asc); + m03710s (w0, w1, w2, w3, pw_len, pws, rules_buf, combs_buf, bfs_buf, tmps, hooks, bitmaps_buf_s1_a, bitmaps_buf_s1_b, bitmaps_buf_s1_c, bitmaps_buf_s1_d, bitmaps_buf_s2_a, bitmaps_buf_s2_b, bitmaps_buf_s2_c, bitmaps_buf_s2_d, plains_buf, digests_buf, hashes_shown, salt_bufs, esalt_bufs, d_return_buf, d_scryptV0_buf, d_scryptV1_buf, d_scryptV2_buf, d_scryptV3_buf, bitmap_mask, bitmap_shift1, bitmap_shift2, salt_pos, loop_pos, loop_cnt, il_cnt, digests_cnt, digests_offset, l_bin2asc); } -__kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m03710_s08 (__global pw_t *pws, __global gpu_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 bfs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max) +__kernel void m03710_s08 (__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_scryptV0_buf, __global u32 *d_scryptV1_buf, __global u32 *d_scryptV2_buf, __global u32 *d_scryptV3_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 */ const u32 gid = get_global_id (0); + const u32 lid = get_local_id (0); + const u32 lsz = get_local_size (0); /** * modifier */ - const u32 lid = get_local_id (0); - u32 w0[4]; w0[0] = pws[gid].i[ 0]; @@ -984,30 +948,14 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m03710_s08 (__glo __local u32 l_bin2asc[256]; - const u32 lid4 = lid * 4; - - const u32 lid40 = lid4 + 0; - const u32 lid41 = lid4 + 1; - const u32 lid42 = lid4 + 2; - const u32 lid43 = lid4 + 3; - - const u32 v400 = (lid40 >> 0) & 15; - const u32 v401 = (lid40 >> 4) & 15; - const u32 v410 = (lid41 >> 0) & 15; - const u32 v411 = (lid41 >> 4) & 15; - const u32 v420 = (lid42 >> 0) & 15; - const u32 v421 = (lid42 >> 4) & 15; - const u32 v430 = (lid43 >> 0) & 15; - const u32 v431 = (lid43 >> 4) & 15; - - l_bin2asc[lid40] = ((v400 < 10) ? '0' + v400 : 'a' - 10 + v400) << 8 - | ((v401 < 10) ? '0' + v401 : 'a' - 10 + v401) << 0; - l_bin2asc[lid41] = ((v410 < 10) ? '0' + v410 : 'a' - 10 + v410) << 8 - | ((v411 < 10) ? '0' + v411 : 'a' - 10 + v411) << 0; - l_bin2asc[lid42] = ((v420 < 10) ? '0' + v420 : 'a' - 10 + v420) << 8 - | ((v421 < 10) ? '0' + v421 : 'a' - 10 + v421) << 0; - l_bin2asc[lid43] = ((v430 < 10) ? '0' + v430 : 'a' - 10 + v430) << 8 - | ((v431 < 10) ? '0' + v431 : 'a' - 10 + v431) << 0; + for (u32 i = lid; i < 256; i += lsz) + { + const u32 i0 = (i >> 0) & 15; + const u32 i1 = (i >> 4) & 15; + + l_bin2asc[i] = ((i0 < 10) ? '0' + i0 : 'a' - 10 + i0) << 8 + | ((i1 < 10) ? '0' + i1 : 'a' - 10 + i1) << 0; + } barrier (CLK_LOCAL_MEM_FENCE); @@ -1017,22 +965,41 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m03710_s08 (__glo * main */ - m03710s (w0, w1, w2, w3, pw_len, pws, rules_buf, combs_buf, bfs_buf, tmps, hooks, bitmaps_buf_s1_a, bitmaps_buf_s1_b, bitmaps_buf_s1_c, bitmaps_buf_s1_d, bitmaps_buf_s2_a, bitmaps_buf_s2_b, bitmaps_buf_s2_c, bitmaps_buf_s2_d, plains_buf, digests_buf, hashes_shown, salt_bufs, esalt_bufs, d_return_buf, d_scryptV_buf, bitmap_mask, bitmap_shift1, bitmap_shift2, salt_pos, loop_pos, loop_cnt, bfs_cnt, digests_cnt, digests_offset, l_bin2asc); + m03710s (w0, w1, w2, w3, pw_len, pws, rules_buf, combs_buf, bfs_buf, tmps, hooks, bitmaps_buf_s1_a, bitmaps_buf_s1_b, bitmaps_buf_s1_c, bitmaps_buf_s1_d, bitmaps_buf_s2_a, bitmaps_buf_s2_b, bitmaps_buf_s2_c, bitmaps_buf_s2_d, plains_buf, digests_buf, hashes_shown, salt_bufs, esalt_bufs, d_return_buf, d_scryptV0_buf, d_scryptV1_buf, d_scryptV2_buf, d_scryptV3_buf, bitmap_mask, bitmap_shift1, bitmap_shift2, salt_pos, loop_pos, loop_cnt, il_cnt, digests_cnt, digests_offset, l_bin2asc); } -__kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m03710_s16 (__global pw_t *pws, __global gpu_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 bfs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max) +__kernel void m03710_s16 (__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_scryptV0_buf, __global u32 *d_scryptV1_buf, __global u32 *d_scryptV2_buf, __global u32 *d_scryptV3_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 */ const u32 gid = get_global_id (0); + const u32 lid = get_local_id (0); + const u32 lsz = get_local_size (0); /** - * modifier + * bin2asc table */ - const u32 lid = get_local_id (0); + __local u32 l_bin2asc[256]; + + for (u32 i = lid; i < 256; i += lsz) + { + const u32 i0 = (i >> 0) & 15; + const u32 i1 = (i >> 4) & 15; + + l_bin2asc[i] = ((i0 < 10) ? '0' + i0 : 'a' - 10 + i0) << 8 + | ((i1 < 10) ? '0' + i1 : 'a' - 10 + i1) << 0; + } + + barrier (CLK_LOCAL_MEM_FENCE); + + if (gid >= gid_max) return; + + /** + * modifier + */ u32 w0[4]; @@ -1064,44 +1031,9 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m03710_s16 (__glo const u32 pw_len = pws[gid].pw_len; - /** - * bin2asc table - */ - - __local u32 l_bin2asc[256]; - - const u32 lid4 = lid * 4; - - const u32 lid40 = lid4 + 0; - const u32 lid41 = lid4 + 1; - const u32 lid42 = lid4 + 2; - const u32 lid43 = lid4 + 3; - - const u32 v400 = (lid40 >> 0) & 15; - const u32 v401 = (lid40 >> 4) & 15; - const u32 v410 = (lid41 >> 0) & 15; - const u32 v411 = (lid41 >> 4) & 15; - const u32 v420 = (lid42 >> 0) & 15; - const u32 v421 = (lid42 >> 4) & 15; - const u32 v430 = (lid43 >> 0) & 15; - const u32 v431 = (lid43 >> 4) & 15; - - l_bin2asc[lid40] = ((v400 < 10) ? '0' + v400 : 'a' - 10 + v400) << 8 - | ((v401 < 10) ? '0' + v401 : 'a' - 10 + v401) << 0; - l_bin2asc[lid41] = ((v410 < 10) ? '0' + v410 : 'a' - 10 + v410) << 8 - | ((v411 < 10) ? '0' + v411 : 'a' - 10 + v411) << 0; - l_bin2asc[lid42] = ((v420 < 10) ? '0' + v420 : 'a' - 10 + v420) << 8 - | ((v421 < 10) ? '0' + v421 : 'a' - 10 + v421) << 0; - l_bin2asc[lid43] = ((v430 < 10) ? '0' + v430 : 'a' - 10 + v430) << 8 - | ((v431 < 10) ? '0' + v431 : 'a' - 10 + v431) << 0; - - barrier (CLK_LOCAL_MEM_FENCE); - - if (gid >= gid_max) return; - /** * main */ - m03710s (w0, w1, w2, w3, pw_len, pws, rules_buf, combs_buf, bfs_buf, tmps, hooks, bitmaps_buf_s1_a, bitmaps_buf_s1_b, bitmaps_buf_s1_c, bitmaps_buf_s1_d, bitmaps_buf_s2_a, bitmaps_buf_s2_b, bitmaps_buf_s2_c, bitmaps_buf_s2_d, plains_buf, digests_buf, hashes_shown, salt_bufs, esalt_bufs, d_return_buf, d_scryptV_buf, bitmap_mask, bitmap_shift1, bitmap_shift2, salt_pos, loop_pos, loop_cnt, bfs_cnt, digests_cnt, digests_offset, l_bin2asc); + m03710s (w0, w1, w2, w3, pw_len, pws, rules_buf, combs_buf, bfs_buf, tmps, hooks, bitmaps_buf_s1_a, bitmaps_buf_s1_b, bitmaps_buf_s1_c, bitmaps_buf_s1_d, bitmaps_buf_s2_a, bitmaps_buf_s2_b, bitmaps_buf_s2_c, bitmaps_buf_s2_d, plains_buf, digests_buf, hashes_shown, salt_bufs, esalt_bufs, d_return_buf, d_scryptV0_buf, d_scryptV1_buf, d_scryptV2_buf, d_scryptV3_buf, bitmap_mask, bitmap_shift1, bitmap_shift2, salt_pos, loop_pos, loop_cnt, il_cnt, digests_cnt, digests_offset, l_bin2asc); }