From 4e97a4db246c3b5bd746effd7613858d2abbb081 Mon Sep 17 00:00:00 2001 From: jsteube Date: Thu, 20 Jul 2017 17:38:43 +0200 Subject: [PATCH] Add pure kernels for md5(sha1($pass)) --- OpenCL/m04400_a0.cl | 253 ++++++++++++++++++++++++++++++++++++++++++ OpenCL/m04400_a1.cl | 229 ++++++++++++++++++++++++++++++++++++++ OpenCL/m04400_a3.cl | 263 ++++++++++++++++++++++++++++++++++++++++++++ 3 files changed, 745 insertions(+) create mode 100644 OpenCL/m04400_a0.cl create mode 100644 OpenCL/m04400_a1.cl create mode 100644 OpenCL/m04400_a3.cl diff --git a/OpenCL/m04400_a0.cl b/OpenCL/m04400_a0.cl new file mode 100644 index 000000000..877fee359 --- /dev/null +++ b/OpenCL/m04400_a0.cl @@ -0,0 +1,253 @@ +/** + * Author......: See docs/credits.txt + * License.....: MIT + */ + +//#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_rp.h" +#include "inc_rp.cl" +#include "inc_scalar.cl" +#include "inc_hash_md5.cl" +#include "inc_hash_sha1.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 + +__kernel void m04400_mxx (__global pw_t *pws, __global const kernel_rule_t *rules_buf, __global const pw_t *combs_buf, __global const bf_t *bfs_buf, __global void *tmps, __global void *hooks, __global const u32 *bitmaps_buf_s1_a, __global const u32 *bitmaps_buf_s1_b, __global const u32 *bitmaps_buf_s1_c, __global const u32 *bitmaps_buf_s1_d, __global const u32 *bitmaps_buf_s2_a, __global const u32 *bitmaps_buf_s2_b, __global const u32 *bitmaps_buf_s2_c, __global const u32 *bitmaps_buf_s2_d, __global plain_t *plains_buf, __global const digest_t *digests_buf, __global u32 *hashes_shown, __global const salt_t *salt_bufs, __global const 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) +{ + /** + * modifier + */ + + const u32 gid = get_global_id (0); + const u32 lid = get_local_id (0); + const u32 lsz = get_local_size (0); + + /** + * bin2asc table + */ + + __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; + + /** + * base + */ + + const u32 pw_len = pws[gid].pw_len; + + const u32 pw_lenv = ceil ((float) pw_len / 4); + + u32 w[64] = { 0 }; + + for (int idx = 0; idx < pw_lenv; idx++) + { + w[idx] = pws[gid].i[idx]; + + barrier (CLK_GLOBAL_MEM_FENCE); + } + + /** + * loop + */ + + for (u32 il_pos = 0; il_pos < il_cnt; il_pos++) + { + // todo: add rules engine + + sha1_ctx_t ctx0; + + sha1_init (&ctx0); + + sha1_update_swap (&ctx0, w, pw_len); + + sha1_final (&ctx0); + + const u32 a = ctx0.h[0]; + const u32 b = ctx0.h[1]; + const u32 c = ctx0.h[2]; + const u32 d = ctx0.h[3]; + const u32 e = ctx0.h[4]; + + md5_ctx_t ctx; + + md5_init (&ctx); + + ctx.w0[0] = uint_to_hex_lower8 ((a >> 24) & 255) << 0 + | uint_to_hex_lower8 ((a >> 16) & 255) << 16; + ctx.w0[1] = uint_to_hex_lower8 ((a >> 8) & 255) << 0 + | uint_to_hex_lower8 ((a >> 0) & 255) << 16; + ctx.w0[2] = uint_to_hex_lower8 ((b >> 24) & 255) << 0 + | uint_to_hex_lower8 ((b >> 16) & 255) << 16; + ctx.w0[3] = uint_to_hex_lower8 ((b >> 8) & 255) << 0 + | uint_to_hex_lower8 ((b >> 0) & 255) << 16; + ctx.w1[0] = uint_to_hex_lower8 ((c >> 24) & 255) << 0 + | uint_to_hex_lower8 ((c >> 16) & 255) << 16; + ctx.w1[1] = uint_to_hex_lower8 ((c >> 8) & 255) << 0 + | uint_to_hex_lower8 ((c >> 0) & 255) << 16; + ctx.w1[2] = uint_to_hex_lower8 ((d >> 24) & 255) << 0 + | uint_to_hex_lower8 ((d >> 16) & 255) << 16; + ctx.w1[3] = uint_to_hex_lower8 ((d >> 8) & 255) << 0 + | uint_to_hex_lower8 ((d >> 0) & 255) << 16; + ctx.w2[0] = uint_to_hex_lower8 ((e >> 24) & 255) << 0 + | uint_to_hex_lower8 ((e >> 16) & 255) << 16; + ctx.w2[1] = uint_to_hex_lower8 ((e >> 8) & 255) << 0 + | uint_to_hex_lower8 ((e >> 0) & 255) << 16; + + ctx.len = 40; + + md5_final (&ctx); + + const u32 r0 = ctx.h[DGST_R0]; + const u32 r1 = ctx.h[DGST_R1]; + const u32 r2 = ctx.h[DGST_R2]; + const u32 r3 = ctx.h[DGST_R3]; + + COMPARE_M_SCALAR (r0, r1, r2, r3); + } +} + +__kernel void m04400_sxx (__global pw_t *pws, __global const kernel_rule_t *rules_buf, __global const pw_t *combs_buf, __global const bf_t *bfs_buf, __global void *tmps, __global void *hooks, __global const u32 *bitmaps_buf_s1_a, __global const u32 *bitmaps_buf_s1_b, __global const u32 *bitmaps_buf_s1_c, __global const u32 *bitmaps_buf_s1_d, __global const u32 *bitmaps_buf_s2_a, __global const u32 *bitmaps_buf_s2_b, __global const u32 *bitmaps_buf_s2_c, __global const u32 *bitmaps_buf_s2_d, __global plain_t *plains_buf, __global const digest_t *digests_buf, __global u32 *hashes_shown, __global const salt_t *salt_bufs, __global const 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) +{ + /** + * modifier + */ + + const u32 gid = get_global_id (0); + const u32 lid = get_local_id (0); + const u32 lsz = get_local_size (0); + + /** + * bin2asc table + */ + + __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; + + /** + * 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] + }; + + /** + * base + */ + + const u32 pw_len = pws[gid].pw_len; + + const u32 pw_lenv = ceil ((float) pw_len / 4); + + u32 w[64] = { 0 }; + + for (int idx = 0; idx < pw_lenv; idx++) + { + w[idx] = pws[gid].i[idx]; + + barrier (CLK_GLOBAL_MEM_FENCE); + } + + /** + * loop + */ + + for (u32 il_pos = 0; il_pos < il_cnt; il_pos++) + { + // todo: add rules engine + + sha1_ctx_t ctx0; + + sha1_init (&ctx0); + + sha1_update_swap (&ctx0, w, pw_len); + + sha1_final (&ctx0); + + const u32 a = ctx0.h[0]; + const u32 b = ctx0.h[1]; + const u32 c = ctx0.h[2]; + const u32 d = ctx0.h[3]; + const u32 e = ctx0.h[4]; + + md5_ctx_t ctx; + + md5_init (&ctx); + + ctx.w0[0] = uint_to_hex_lower8 ((a >> 24) & 255) << 0 + | uint_to_hex_lower8 ((a >> 16) & 255) << 16; + ctx.w0[1] = uint_to_hex_lower8 ((a >> 8) & 255) << 0 + | uint_to_hex_lower8 ((a >> 0) & 255) << 16; + ctx.w0[2] = uint_to_hex_lower8 ((b >> 24) & 255) << 0 + | uint_to_hex_lower8 ((b >> 16) & 255) << 16; + ctx.w0[3] = uint_to_hex_lower8 ((b >> 8) & 255) << 0 + | uint_to_hex_lower8 ((b >> 0) & 255) << 16; + ctx.w1[0] = uint_to_hex_lower8 ((c >> 24) & 255) << 0 + | uint_to_hex_lower8 ((c >> 16) & 255) << 16; + ctx.w1[1] = uint_to_hex_lower8 ((c >> 8) & 255) << 0 + | uint_to_hex_lower8 ((c >> 0) & 255) << 16; + ctx.w1[2] = uint_to_hex_lower8 ((d >> 24) & 255) << 0 + | uint_to_hex_lower8 ((d >> 16) & 255) << 16; + ctx.w1[3] = uint_to_hex_lower8 ((d >> 8) & 255) << 0 + | uint_to_hex_lower8 ((d >> 0) & 255) << 16; + ctx.w2[0] = uint_to_hex_lower8 ((e >> 24) & 255) << 0 + | uint_to_hex_lower8 ((e >> 16) & 255) << 16; + ctx.w2[1] = uint_to_hex_lower8 ((e >> 8) & 255) << 0 + | uint_to_hex_lower8 ((e >> 0) & 255) << 16; + + ctx.len = 40; + + md5_final (&ctx); + + const u32 r0 = ctx.h[DGST_R0]; + const u32 r1 = ctx.h[DGST_R1]; + const u32 r2 = ctx.h[DGST_R2]; + const u32 r3 = ctx.h[DGST_R3]; + + COMPARE_S_SCALAR (r0, r1, r2, r3); + } +} diff --git a/OpenCL/m04400_a1.cl b/OpenCL/m04400_a1.cl new file mode 100644 index 000000000..190138933 --- /dev/null +++ b/OpenCL/m04400_a1.cl @@ -0,0 +1,229 @@ +/** + * Author......: See docs/credits.txt + * License.....: MIT + */ + +//#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_scalar.cl" +#include "inc_hash_md5.cl" +#include "inc_hash_sha1.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 + +__kernel void m04400_mxx (__global pw_t *pws, __global const kernel_rule_t *rules_buf, __global const pw_t *combs_buf, __global const bf_t *bfs_buf, __global void *tmps, __global void *hooks, __global const u32 *bitmaps_buf_s1_a, __global const u32 *bitmaps_buf_s1_b, __global const u32 *bitmaps_buf_s1_c, __global const u32 *bitmaps_buf_s1_d, __global const u32 *bitmaps_buf_s2_a, __global const u32 *bitmaps_buf_s2_b, __global const u32 *bitmaps_buf_s2_c, __global const u32 *bitmaps_buf_s2_d, __global plain_t *plains_buf, __global const digest_t *digests_buf, __global u32 *hashes_shown, __global const salt_t *salt_bufs, __global const 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) +{ + /** + * modifier + */ + + const u32 gid = get_global_id (0); + const u32 lid = get_local_id (0); + const u32 lsz = get_local_size (0); + + /** + * bin2asc table + */ + + __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; + + /** + * base + */ + + sha1_ctx_t ctx0; + + sha1_init (&ctx0); + + sha1_update_global_swap (&ctx0, pws[gid].i, pws[gid].pw_len); + + /** + * loop + */ + + for (u32 il_pos = 0; il_pos < il_cnt; il_pos++) + { + sha1_ctx_t ctx1 = ctx0; + + sha1_update_global_swap (&ctx1, combs_buf[il_pos].i, combs_buf[il_pos].pw_len); + + sha1_final (&ctx1); + + const u32 a = ctx1.h[0]; + const u32 b = ctx1.h[1]; + const u32 c = ctx1.h[2]; + const u32 d = ctx1.h[3]; + const u32 e = ctx1.h[4]; + + md5_ctx_t ctx; + + md5_init (&ctx); + + ctx.w0[0] = uint_to_hex_lower8 ((a >> 24) & 255) << 0 + | uint_to_hex_lower8 ((a >> 16) & 255) << 16; + ctx.w0[1] = uint_to_hex_lower8 ((a >> 8) & 255) << 0 + | uint_to_hex_lower8 ((a >> 0) & 255) << 16; + ctx.w0[2] = uint_to_hex_lower8 ((b >> 24) & 255) << 0 + | uint_to_hex_lower8 ((b >> 16) & 255) << 16; + ctx.w0[3] = uint_to_hex_lower8 ((b >> 8) & 255) << 0 + | uint_to_hex_lower8 ((b >> 0) & 255) << 16; + ctx.w1[0] = uint_to_hex_lower8 ((c >> 24) & 255) << 0 + | uint_to_hex_lower8 ((c >> 16) & 255) << 16; + ctx.w1[1] = uint_to_hex_lower8 ((c >> 8) & 255) << 0 + | uint_to_hex_lower8 ((c >> 0) & 255) << 16; + ctx.w1[2] = uint_to_hex_lower8 ((d >> 24) & 255) << 0 + | uint_to_hex_lower8 ((d >> 16) & 255) << 16; + ctx.w1[3] = uint_to_hex_lower8 ((d >> 8) & 255) << 0 + | uint_to_hex_lower8 ((d >> 0) & 255) << 16; + ctx.w2[0] = uint_to_hex_lower8 ((e >> 24) & 255) << 0 + | uint_to_hex_lower8 ((e >> 16) & 255) << 16; + ctx.w2[1] = uint_to_hex_lower8 ((e >> 8) & 255) << 0 + | uint_to_hex_lower8 ((e >> 0) & 255) << 16; + + ctx.len = 40; + + md5_final (&ctx); + + const u32 r0 = ctx.h[DGST_R0]; + const u32 r1 = ctx.h[DGST_R1]; + const u32 r2 = ctx.h[DGST_R2]; + const u32 r3 = ctx.h[DGST_R3]; + + COMPARE_M_SCALAR (r0, r1, r2, r3); + } +} + +__kernel void m04400_sxx (__global pw_t *pws, __global const kernel_rule_t *rules_buf, __global const pw_t *combs_buf, __global const bf_t *bfs_buf, __global void *tmps, __global void *hooks, __global const u32 *bitmaps_buf_s1_a, __global const u32 *bitmaps_buf_s1_b, __global const u32 *bitmaps_buf_s1_c, __global const u32 *bitmaps_buf_s1_d, __global const u32 *bitmaps_buf_s2_a, __global const u32 *bitmaps_buf_s2_b, __global const u32 *bitmaps_buf_s2_c, __global const u32 *bitmaps_buf_s2_d, __global plain_t *plains_buf, __global const digest_t *digests_buf, __global u32 *hashes_shown, __global const salt_t *salt_bufs, __global const 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) +{ + /** + * modifier + */ + + const u32 gid = get_global_id (0); + const u32 lid = get_local_id (0); + const u32 lsz = get_local_size (0); + + /** + * bin2asc table + */ + + __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; + + /** + * 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] + }; + + /** + * base + */ + + sha1_ctx_t ctx0; + + sha1_init (&ctx0); + + sha1_update_global_swap (&ctx0, pws[gid].i, pws[gid].pw_len); + + /** + * loop + */ + + for (u32 il_pos = 0; il_pos < il_cnt; il_pos++) + { + sha1_ctx_t ctx1 = ctx0; + + sha1_update_global_swap (&ctx1, combs_buf[il_pos].i, combs_buf[il_pos].pw_len); + + sha1_final (&ctx1); + + const u32 a = ctx1.h[0]; + const u32 b = ctx1.h[1]; + const u32 c = ctx1.h[2]; + const u32 d = ctx1.h[3]; + const u32 e = ctx1.h[4]; + + md5_ctx_t ctx; + + md5_init (&ctx); + + ctx.w0[0] = uint_to_hex_lower8 ((a >> 24) & 255) << 0 + | uint_to_hex_lower8 ((a >> 16) & 255) << 16; + ctx.w0[1] = uint_to_hex_lower8 ((a >> 8) & 255) << 0 + | uint_to_hex_lower8 ((a >> 0) & 255) << 16; + ctx.w0[2] = uint_to_hex_lower8 ((b >> 24) & 255) << 0 + | uint_to_hex_lower8 ((b >> 16) & 255) << 16; + ctx.w0[3] = uint_to_hex_lower8 ((b >> 8) & 255) << 0 + | uint_to_hex_lower8 ((b >> 0) & 255) << 16; + ctx.w1[0] = uint_to_hex_lower8 ((c >> 24) & 255) << 0 + | uint_to_hex_lower8 ((c >> 16) & 255) << 16; + ctx.w1[1] = uint_to_hex_lower8 ((c >> 8) & 255) << 0 + | uint_to_hex_lower8 ((c >> 0) & 255) << 16; + ctx.w1[2] = uint_to_hex_lower8 ((d >> 24) & 255) << 0 + | uint_to_hex_lower8 ((d >> 16) & 255) << 16; + ctx.w1[3] = uint_to_hex_lower8 ((d >> 8) & 255) << 0 + | uint_to_hex_lower8 ((d >> 0) & 255) << 16; + ctx.w2[0] = uint_to_hex_lower8 ((e >> 24) & 255) << 0 + | uint_to_hex_lower8 ((e >> 16) & 255) << 16; + ctx.w2[1] = uint_to_hex_lower8 ((e >> 8) & 255) << 0 + | uint_to_hex_lower8 ((e >> 0) & 255) << 16; + + ctx.len = 40; + + md5_final (&ctx); + + const u32 r0 = ctx.h[DGST_R0]; + const u32 r1 = ctx.h[DGST_R1]; + const u32 r2 = ctx.h[DGST_R2]; + const u32 r3 = ctx.h[DGST_R3]; + + COMPARE_S_SCALAR (r0, r1, r2, r3); + } +} diff --git a/OpenCL/m04400_a3.cl b/OpenCL/m04400_a3.cl new file mode 100644 index 000000000..598de7915 --- /dev/null +++ b/OpenCL/m04400_a3.cl @@ -0,0 +1,263 @@ +/** + * Author......: See docs/credits.txt + * License.....: MIT + */ + +#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" +#include "inc_hash_md5.cl" +#include "inc_hash_sha1.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 + +__kernel void m04400_mxx (__global pw_t *pws, __global const kernel_rule_t *rules_buf, __global const pw_t *combs_buf, __constant const u32x *words_buf_r, __global void *tmps, __global void *hooks, __global const u32 *bitmaps_buf_s1_a, __global const u32 *bitmaps_buf_s1_b, __global const u32 *bitmaps_buf_s1_c, __global const u32 *bitmaps_buf_s1_d, __global const u32 *bitmaps_buf_s2_a, __global const u32 *bitmaps_buf_s2_b, __global const u32 *bitmaps_buf_s2_c, __global const u32 *bitmaps_buf_s2_d, __global plain_t *plains_buf, __global const digest_t *digests_buf, __global u32 *hashes_shown, __global const salt_t *salt_bufs, __global const 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) +{ + /** + * modifier + */ + + const u32 gid = get_global_id (0); + const u32 lid = get_local_id (0); + const u32 lsz = get_local_size (0); + + /** + * bin2asc table + */ + + __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; + + /** + * base + */ + + const u32 pw_len = pws[gid].pw_len; + + const u32 pw_lenv = ceil ((float) pw_len / 4); + + u32x w[64] = { 0 }; + + for (int idx = 0; idx < pw_lenv; idx++) + { + w[idx] = pws[gid].i[idx]; + + barrier (CLK_GLOBAL_MEM_FENCE); + } + + /** + * loop + */ + + u32x w0l = w[0]; + + 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 w0 = w0l | w0r; + + w[0] = w0; + + sha1_ctx_vector_t ctx0; + + sha1_init_vector (&ctx0); + + sha1_update_vector (&ctx0, w, pw_len); + + sha1_final_vector (&ctx0); + + const u32x a = ctx0.h[0]; + const u32x b = ctx0.h[1]; + const u32x c = ctx0.h[2]; + const u32x d = ctx0.h[3]; + const u32x e = ctx0.h[4]; + + md5_ctx_vector_t ctx; + + md5_init_vector (&ctx); + + ctx.w0[0] = uint_to_hex_lower8 ((a >> 24) & 255) << 0 + | uint_to_hex_lower8 ((a >> 16) & 255) << 16; + ctx.w0[1] = uint_to_hex_lower8 ((a >> 8) & 255) << 0 + | uint_to_hex_lower8 ((a >> 0) & 255) << 16; + ctx.w0[2] = uint_to_hex_lower8 ((b >> 24) & 255) << 0 + | uint_to_hex_lower8 ((b >> 16) & 255) << 16; + ctx.w0[3] = uint_to_hex_lower8 ((b >> 8) & 255) << 0 + | uint_to_hex_lower8 ((b >> 0) & 255) << 16; + ctx.w1[0] = uint_to_hex_lower8 ((c >> 24) & 255) << 0 + | uint_to_hex_lower8 ((c >> 16) & 255) << 16; + ctx.w1[1] = uint_to_hex_lower8 ((c >> 8) & 255) << 0 + | uint_to_hex_lower8 ((c >> 0) & 255) << 16; + ctx.w1[2] = uint_to_hex_lower8 ((d >> 24) & 255) << 0 + | uint_to_hex_lower8 ((d >> 16) & 255) << 16; + ctx.w1[3] = uint_to_hex_lower8 ((d >> 8) & 255) << 0 + | uint_to_hex_lower8 ((d >> 0) & 255) << 16; + ctx.w2[0] = uint_to_hex_lower8 ((e >> 24) & 255) << 0 + | uint_to_hex_lower8 ((e >> 16) & 255) << 16; + ctx.w2[1] = uint_to_hex_lower8 ((e >> 8) & 255) << 0 + | uint_to_hex_lower8 ((e >> 0) & 255) << 16; + + ctx.len = 40; + + md5_final_vector (&ctx); + + const u32x r0 = ctx.h[DGST_R0]; + const u32x r1 = ctx.h[DGST_R1]; + const u32x r2 = ctx.h[DGST_R2]; + const u32x r3 = ctx.h[DGST_R3]; + + COMPARE_M_SIMD (r0, r1, r2, r3); + } +} + +__kernel void m04400_sxx (__global pw_t *pws, __global const kernel_rule_t *rules_buf, __global const pw_t *combs_buf, __constant const u32x *words_buf_r, __global void *tmps, __global void *hooks, __global const u32 *bitmaps_buf_s1_a, __global const u32 *bitmaps_buf_s1_b, __global const u32 *bitmaps_buf_s1_c, __global const u32 *bitmaps_buf_s1_d, __global const u32 *bitmaps_buf_s2_a, __global const u32 *bitmaps_buf_s2_b, __global const u32 *bitmaps_buf_s2_c, __global const u32 *bitmaps_buf_s2_d, __global plain_t *plains_buf, __global const digest_t *digests_buf, __global u32 *hashes_shown, __global const salt_t *salt_bufs, __global const 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) +{ + /** + * modifier + */ + + const u32 gid = get_global_id (0); + const u32 lid = get_local_id (0); + const u32 lsz = get_local_size (0); + + /** + * bin2asc table + */ + + __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; + + /** + * 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] + }; + + /** + * base + */ + + const u32 pw_len = pws[gid].pw_len; + + const u32 pw_lenv = ceil ((float) pw_len / 4); + + u32x w[64] = { 0 }; + + for (int idx = 0; idx < pw_lenv; idx++) + { + w[idx] = pws[gid].i[idx]; + + barrier (CLK_GLOBAL_MEM_FENCE); + } + + /** + * loop + */ + + u32x w0l = w[0]; + + 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 w0 = w0l | w0r; + + w[0] = w0; + + sha1_ctx_vector_t ctx0; + + sha1_init_vector (&ctx0); + + sha1_update_vector (&ctx0, w, pw_len); + + sha1_final_vector (&ctx0); + + const u32x a = ctx0.h[0]; + const u32x b = ctx0.h[1]; + const u32x c = ctx0.h[2]; + const u32x d = ctx0.h[3]; + const u32x e = ctx0.h[4]; + + md5_ctx_vector_t ctx; + + md5_init_vector (&ctx); + + ctx.w0[0] = uint_to_hex_lower8 ((a >> 24) & 255) << 0 + | uint_to_hex_lower8 ((a >> 16) & 255) << 16; + ctx.w0[1] = uint_to_hex_lower8 ((a >> 8) & 255) << 0 + | uint_to_hex_lower8 ((a >> 0) & 255) << 16; + ctx.w0[2] = uint_to_hex_lower8 ((b >> 24) & 255) << 0 + | uint_to_hex_lower8 ((b >> 16) & 255) << 16; + ctx.w0[3] = uint_to_hex_lower8 ((b >> 8) & 255) << 0 + | uint_to_hex_lower8 ((b >> 0) & 255) << 16; + ctx.w1[0] = uint_to_hex_lower8 ((c >> 24) & 255) << 0 + | uint_to_hex_lower8 ((c >> 16) & 255) << 16; + ctx.w1[1] = uint_to_hex_lower8 ((c >> 8) & 255) << 0 + | uint_to_hex_lower8 ((c >> 0) & 255) << 16; + ctx.w1[2] = uint_to_hex_lower8 ((d >> 24) & 255) << 0 + | uint_to_hex_lower8 ((d >> 16) & 255) << 16; + ctx.w1[3] = uint_to_hex_lower8 ((d >> 8) & 255) << 0 + | uint_to_hex_lower8 ((d >> 0) & 255) << 16; + ctx.w2[0] = uint_to_hex_lower8 ((e >> 24) & 255) << 0 + | uint_to_hex_lower8 ((e >> 16) & 255) << 16; + ctx.w2[1] = uint_to_hex_lower8 ((e >> 8) & 255) << 0 + | uint_to_hex_lower8 ((e >> 0) & 255) << 16; + + ctx.len = 40; + + md5_final_vector (&ctx); + + const u32x r0 = ctx.h[DGST_R0]; + const u32x r1 = ctx.h[DGST_R1]; + const u32x r2 = ctx.h[DGST_R2]; + const u32x r3 = ctx.h[DGST_R3]; + + COMPARE_S_SIMD (r0, r1, r2, r3); + } +}