X-Git-Url: https://www.flypig.org.uk/git/?a=blobdiff_plain;f=OpenCL%2Fm05400_a3.cl;h=4d26407d291f62ce7c2410b5bf0e9fc597309164;hb=51e1d11b2d6dff4ae331cff6098169455733c412;hp=215d98b14f78aeeab7c461a10f4508e09616f561;hpb=6893a9ad857a1d4fe63f1a92365802f5d9d3b39e;p=hashcat.git diff --git a/OpenCL/m05400_a3.cl b/OpenCL/m05400_a3.cl index 215d98b..4d26407 100644 --- a/OpenCL/m05400_a3.cl +++ b/OpenCL/m05400_a3.cl @@ -1,49 +1,45 @@ /** - * Author......: Jens Steube + * Authors.....: Jens Steube + * Gabriele Gristina + * * License.....: MIT */ #define _SHA1_ -#include "include/constants.h" -#include "include/kernel_vendor.h" +#define NEW_SIMD_CODE -#define DGST_R0 3 -#define DGST_R1 4 -#define DGST_R2 2 -#define DGST_R3 1 +#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 "include/kernel_functions.c" -#include "types_ocl.c" -#include "common.c" - -#define COMPARE_S "check_single_comp4.c" -#define COMPARE_M "check_multi_comp4.c" - -static void sha1_transform (const u32 w0[4], const u32 w1[4], const u32 w2[4], const u32 w3[4], u32 digest[5]) +void sha1_transform (const u32x w0[4], const u32x w1[4], const u32x w2[4], const u32x w3[4], u32x digest[5]) { - u32 A = digest[0]; - u32 B = digest[1]; - u32 C = digest[2]; - u32 D = digest[3]; - u32 E = digest[4]; - - u32 w0_t = w0[0]; - u32 w1_t = w0[1]; - u32 w2_t = w0[2]; - u32 w3_t = w0[3]; - u32 w4_t = w1[0]; - u32 w5_t = w1[1]; - u32 w6_t = w1[2]; - u32 w7_t = w1[3]; - u32 w8_t = w2[0]; - u32 w9_t = w2[1]; - u32 wa_t = w2[2]; - u32 wb_t = w2[3]; - u32 wc_t = w3[0]; - u32 wd_t = w3[1]; - u32 we_t = w3[2]; - u32 wf_t = w3[3]; + u32x A = digest[0]; + u32x B = digest[1]; + u32x C = digest[2]; + u32x D = digest[3]; + u32x E = digest[4]; + + u32x w0_t = w0[0]; + u32x w1_t = w0[1]; + u32x w2_t = w0[2]; + u32x w3_t = w0[3]; + u32x w4_t = w1[0]; + u32x w5_t = w1[1]; + u32x w6_t = w1[2]; + u32x w7_t = w1[3]; + u32x w8_t = w2[0]; + u32x w9_t = w2[1]; + u32x wa_t = w2[2]; + u32x wb_t = w2[3]; + u32x wc_t = w3[0]; + u32x wd_t = w3[1]; + u32x we_t = w3[2]; + u32x wf_t = w3[3]; #undef K #define K SHA1C00 @@ -148,7 +144,7 @@ static void sha1_transform (const u32 w0[4], const u32 w1[4], const u32 w2[4], c digest[4] += E; } -static void hmac_sha1_pad (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], u32 ipad[5], u32 opad[5]) +void hmac_sha1_pad (u32x w0[4], u32x w1[4], u32x w2[4], u32x w3[4], u32x ipad[5], u32x opad[5]) { w0[0] = w0[0] ^ 0x36363636; w0[1] = w0[1] ^ 0x36363636; @@ -201,7 +197,7 @@ static void hmac_sha1_pad (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], u32 ipad[ sha1_transform (w0, w1, w2, w3, opad); } -static void hmac_sha1_run (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], u32 ipad[5], u32 opad[5], u32 digest[5]) +void hmac_sha1_run (u32x w0[4], u32x w1[4], u32x w2[4], u32x w3[4], u32x ipad[5], u32x opad[5], u32x digest[5]) { digest[0] = ipad[0]; digest[1] = ipad[1]; @@ -237,7 +233,7 @@ static void hmac_sha1_run (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], u32 ipad[ sha1_transform (w0, w1, w2, w3, digest); } -static void m05400m (__local u32 w_s[16], 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 ikepsk_t *ikepsk_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 s_msg_buf[128]) +void m05400m (__local u32 *w_s, 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 ikepsk_t *ikepsk_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 *s_msg_buf) { /** * modifier @@ -259,46 +255,40 @@ static void m05400m (__local u32 w_s[16], u32 w0[4], u32 w1[4], u32 w2[4], u32 w 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; + const u32x w0r = ix_create_bft (bfs_buf, il_pos); - w0[0] = w0l | w0r; - - /** - * pads - */ + const u32x w0lr = w0l | w0r; - u32 w0_t[4]; + u32x w0_t[4]; + u32x w1_t[4]; + u32x w2_t[4]; + u32x w3_t[4]; - w0_t[0] = w0[0]; + w0_t[0] = w0lr; w0_t[1] = w0[1]; w0_t[2] = w0[2]; w0_t[3] = w0[3]; - - u32 w1_t[4]; - w1_t[0] = w1[0]; w1_t[1] = w1[1]; w1_t[2] = w1[2]; w1_t[3] = w1[3]; - - u32 w2_t[4]; - w2_t[0] = w2[0]; w2_t[1] = w2[1]; w2_t[2] = w2[2]; w2_t[3] = w2[3]; - - u32 w3_t[4]; - w3_t[0] = w3[0]; w3_t[1] = w3[1]; - w3_t[2] = 0; - w3_t[3] = 0; + w3_t[2] = w3[2]; + w3_t[3] = w3[3]; + + /** + * pads + */ - u32 ipad[5]; - u32 opad[5]; + u32x ipad[5]; + u32x opad[5]; hmac_sha1_pad (w0_t, w1_t, w2_t, w3_t, ipad, opad); @@ -319,7 +309,7 @@ static void m05400m (__local u32 w_s[16], u32 w0[4], u32 w1[4], u32 w2[4], u32 w w3_t[2] = 0; w3_t[3] = (64 + nr_len) * 8; - u32 digest[5]; + u32x digest[5]; hmac_sha1_run (w0_t, w1_t, w2_t, w3_t, ipad, opad, digest); @@ -386,16 +376,11 @@ static void m05400m (__local u32 w_s[16], u32 w0[4], u32 w1[4], u32 w2[4], u32 w hmac_sha1_run (w0_t, w1_t, w2_t, w3_t, ipad, opad, digest); - const u32 r0 = digest[3]; - const u32 r1 = digest[4]; - const u32 r2 = digest[2]; - const u32 r3 = digest[1]; - - #include COMPARE_M + COMPARE_M_SIMD (digest[3], digest[4], digest[2], digest[1]); } } -static void m05400s (__local u32 w_s[16], 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 ikepsk_t *ikepsk_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 s_msg_buf[128]) +void m05400s (__local u32 *w_s, 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 ikepsk_t *ikepsk_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 *s_msg_buf) { /** * modifier @@ -429,46 +414,40 @@ static void m05400s (__local u32 w_s[16], u32 w0[4], u32 w1[4], u32 w2[4], u32 w 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; + const u32x w0r = ix_create_bft (bfs_buf, il_pos); - w0[0] = w0l | w0r; - - /** - * pads - */ + const u32x w0lr = w0l | w0r; - u32 w0_t[4]; + u32x w0_t[4]; + u32x w1_t[4]; + u32x w2_t[4]; + u32x w3_t[4]; - w0_t[0] = w0[0]; + w0_t[0] = w0lr; w0_t[1] = w0[1]; w0_t[2] = w0[2]; w0_t[3] = w0[3]; - - u32 w1_t[4]; - w1_t[0] = w1[0]; w1_t[1] = w1[1]; w1_t[2] = w1[2]; w1_t[3] = w1[3]; - - u32 w2_t[4]; - w2_t[0] = w2[0]; w2_t[1] = w2[1]; w2_t[2] = w2[2]; w2_t[3] = w2[3]; - - u32 w3_t[4]; - w3_t[0] = w3[0]; w3_t[1] = w3[1]; - w3_t[2] = 0; - w3_t[3] = 0; + w3_t[2] = w3[2]; + w3_t[3] = w3[3]; + + /** + * pads + */ - u32 ipad[5]; - u32 opad[5]; + u32x ipad[5]; + u32x opad[5]; hmac_sha1_pad (w0_t, w1_t, w2_t, w3_t, ipad, opad); @@ -489,7 +468,7 @@ static void m05400s (__local u32 w_s[16], u32 w0[4], u32 w1[4], u32 w2[4], u32 w w3_t[2] = 0; w3_t[3] = (64 + nr_len) * 8; - u32 digest[5]; + u32x digest[5]; hmac_sha1_run (w0_t, w1_t, w2_t, w3_t, ipad, opad, digest); @@ -556,28 +535,47 @@ static void m05400s (__local u32 w_s[16], u32 w0[4], u32 w1[4], u32 w2[4], u32 w hmac_sha1_run (w0_t, w1_t, w2_t, w3_t, ipad, opad, digest); - const u32 r0 = digest[3]; - const u32 r1 = digest[4]; - const u32 r2 = digest[2]; - const u32 r3 = digest[1]; - - #include COMPARE_S + COMPARE_S_SIMD (digest[3], digest[4], digest[2], digest[1]); } } -__kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m05400_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 ikepsk_t *ikepsk_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 m05400_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 ikepsk_t *ikepsk_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); /** - * base + * s_msg */ - const u32 gid = get_global_id (0); + __local u32 w_s[16]; + + for (u32 i = lid; i < 16; i += lsz) + { + w_s[i] = swap32_S (ikepsk_bufs[salt_pos].nr_buf[i]); + } + + barrier (CLK_LOCAL_MEM_FENCE); + + __local u32 s_msg_buf[128]; + + for (u32 i = lid; i < 128; i += lsz) + { + s_msg_buf[i] = swap32_S (ikepsk_bufs[salt_pos].msg_buf[i]); + } + + barrier (CLK_LOCAL_MEM_FENCE); + + if (gid >= gid_max) return; + + /** + * base + */ u32 w0[4]; @@ -609,51 +607,51 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m05400_m04 (__glo const u32 pw_len = pws[gid].pw_len; + /** + * main + */ + + m05400m (w_s, 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, ikepsk_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, s_msg_buf); +} + +__kernel void m05400_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 ikepsk_t *ikepsk_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); + /** * s_msg */ __local u32 w_s[16]; - if (lid < 16) + for (u32 i = lid; i < 16; i += lsz) { - w_s[lid] = swap32 (ikepsk_bufs[salt_pos].nr_buf[lid]); + w_s[i] = swap32_S (ikepsk_bufs[salt_pos].nr_buf[i]); } barrier (CLK_LOCAL_MEM_FENCE); __local u32 s_msg_buf[128]; - const u32 lid2 = lid * 2; - - s_msg_buf[lid2 + 0] = swap32 (ikepsk_bufs[salt_pos].msg_buf[lid2 + 0]); - s_msg_buf[lid2 + 1] = swap32 (ikepsk_bufs[salt_pos].msg_buf[lid2 + 1]); + for (u32 i = lid; i < 128; i += lsz) + { + s_msg_buf[i] = swap32_S (ikepsk_bufs[salt_pos].msg_buf[i]); + } barrier (CLK_LOCAL_MEM_FENCE); if (gid >= gid_max) return; - /** - * main - */ - - m05400m (w_s, 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, ikepsk_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, s_msg_buf); -} - -__kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m05400_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 ikepsk_t *ikepsk_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) -{ - /** - * modifier - */ - - const u32 lid = get_local_id (0); - /** * base */ - const u32 gid = get_global_id (0); - u32 w0[4]; w0[0] = pws[gid].i[ 0]; @@ -684,51 +682,51 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m05400_m08 (__glo const u32 pw_len = pws[gid].pw_len; + /** + * main + */ + + m05400m (w_s, 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, ikepsk_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, s_msg_buf); +} + +__kernel void m05400_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 ikepsk_t *ikepsk_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); + /** * s_msg */ __local u32 w_s[16]; - if (lid < 16) + for (u32 i = lid; i < 16; i += lsz) { - w_s[lid] = swap32 (ikepsk_bufs[salt_pos].nr_buf[lid]); + w_s[i] = swap32_S (ikepsk_bufs[salt_pos].nr_buf[i]); } barrier (CLK_LOCAL_MEM_FENCE); __local u32 s_msg_buf[128]; - const u32 lid2 = lid * 2; - - s_msg_buf[lid2 + 0] = swap32 (ikepsk_bufs[salt_pos].msg_buf[lid2 + 0]); - s_msg_buf[lid2 + 1] = swap32 (ikepsk_bufs[salt_pos].msg_buf[lid2 + 1]); + for (u32 i = lid; i < 128; i += lsz) + { + s_msg_buf[i] = swap32_S (ikepsk_bufs[salt_pos].msg_buf[i]); + } barrier (CLK_LOCAL_MEM_FENCE); if (gid >= gid_max) return; - /** - * main - */ - - m05400m (w_s, 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, ikepsk_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, s_msg_buf); -} - -__kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m05400_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 ikepsk_t *ikepsk_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) -{ - /** - * modifier - */ - - const u32 lid = get_local_id (0); - /** * base */ - const u32 gid = get_global_id (0); - u32 w0[4]; w0[0] = pws[gid].i[ 0]; @@ -759,51 +757,51 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m05400_m16 (__glo const u32 pw_len = pws[gid].pw_len; + /** + * main + */ + + m05400m (w_s, 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, ikepsk_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, s_msg_buf); +} + +__kernel void m05400_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 ikepsk_t *ikepsk_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); + /** * s_msg */ __local u32 w_s[16]; - if (lid < 16) + for (u32 i = lid; i < 16; i += lsz) { - w_s[lid] = swap32 (ikepsk_bufs[salt_pos].nr_buf[lid]); + w_s[i] = swap32_S (ikepsk_bufs[salt_pos].nr_buf[i]); } barrier (CLK_LOCAL_MEM_FENCE); __local u32 s_msg_buf[128]; - const u32 lid2 = lid * 2; - - s_msg_buf[lid2 + 0] = swap32 (ikepsk_bufs[salt_pos].msg_buf[lid2 + 0]); - s_msg_buf[lid2 + 1] = swap32 (ikepsk_bufs[salt_pos].msg_buf[lid2 + 1]); + for (u32 i = lid; i < 128; i += lsz) + { + s_msg_buf[i] = swap32_S (ikepsk_bufs[salt_pos].msg_buf[i]); + } barrier (CLK_LOCAL_MEM_FENCE); if (gid >= gid_max) return; - /** - * main - */ - - m05400m (w_s, 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, ikepsk_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, s_msg_buf); -} - -__kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m05400_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 ikepsk_t *ikepsk_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) -{ - /** - * modifier - */ - - const u32 lid = get_local_id (0); - /** * base */ - const u32 gid = get_global_id (0); - u32 w0[4]; w0[0] = pws[gid].i[ 0]; @@ -834,51 +832,51 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m05400_s04 (__glo const u32 pw_len = pws[gid].pw_len; - /* - * s_msg_buf + /** + * main + */ + + m05400s (w_s, 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, ikepsk_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, s_msg_buf); +} + +__kernel void m05400_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 ikepsk_t *ikepsk_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); + + /** + * s_msg */ __local u32 w_s[16]; - if (lid < 16) + for (u32 i = lid; i < 16; i += lsz) { - w_s[lid] = swap32 (ikepsk_bufs[salt_pos].nr_buf[lid]); + w_s[i] = swap32_S (ikepsk_bufs[salt_pos].nr_buf[i]); } barrier (CLK_LOCAL_MEM_FENCE); __local u32 s_msg_buf[128]; - const u32 lid2 = lid * 2; - - s_msg_buf[lid2 + 0] = swap32 (ikepsk_bufs[salt_pos].msg_buf[lid2 + 0]); - s_msg_buf[lid2 + 1] = swap32 (ikepsk_bufs[salt_pos].msg_buf[lid2 + 1]); + for (u32 i = lid; i < 128; i += lsz) + { + s_msg_buf[i] = swap32_S (ikepsk_bufs[salt_pos].msg_buf[i]); + } barrier (CLK_LOCAL_MEM_FENCE); if (gid >= gid_max) return; - /** - * main - */ - - m05400s (w_s, 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, ikepsk_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, s_msg_buf); -} - -__kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m05400_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 ikepsk_t *ikepsk_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) -{ - /** - * modifier - */ - - const u32 lid = get_local_id (0); - /** * base */ - const u32 gid = get_global_id (0); - u32 w0[4]; w0[0] = pws[gid].i[ 0]; @@ -909,51 +907,51 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m05400_s08 (__glo const u32 pw_len = pws[gid].pw_len; - /* - * s_msg_buf + /** + * main + */ + + m05400s (w_s, 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, ikepsk_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, s_msg_buf); +} + +__kernel void m05400_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 ikepsk_t *ikepsk_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); + + /** + * s_msg */ __local u32 w_s[16]; - if (lid < 16) + for (u32 i = lid; i < 16; i += lsz) { - w_s[lid] = swap32 (ikepsk_bufs[salt_pos].nr_buf[lid]); + w_s[i] = swap32_S (ikepsk_bufs[salt_pos].nr_buf[i]); } barrier (CLK_LOCAL_MEM_FENCE); __local u32 s_msg_buf[128]; - const u32 lid2 = lid * 2; - - s_msg_buf[lid2 + 0] = swap32 (ikepsk_bufs[salt_pos].msg_buf[lid2 + 0]); - s_msg_buf[lid2 + 1] = swap32 (ikepsk_bufs[salt_pos].msg_buf[lid2 + 1]); + for (u32 i = lid; i < 128; i += lsz) + { + s_msg_buf[i] = swap32_S (ikepsk_bufs[salt_pos].msg_buf[i]); + } barrier (CLK_LOCAL_MEM_FENCE); if (gid >= gid_max) return; - /** - * main - */ - - m05400s (w_s, 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, ikepsk_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, s_msg_buf); -} - -__kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m05400_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 ikepsk_t *ikepsk_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) -{ - /** - * modifier - */ - - const u32 lid = get_local_id (0); - /** * base */ - const u32 gid = get_global_id (0); - u32 w0[4]; w0[0] = pws[gid].i[ 0]; @@ -984,33 +982,9 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m05400_s16 (__glo const u32 pw_len = pws[gid].pw_len; - /* - * s_msg_buf - */ - - __local u32 w_s[16]; - - if (lid < 16) - { - w_s[lid] = swap32 (ikepsk_bufs[salt_pos].nr_buf[lid]); - } - - barrier (CLK_LOCAL_MEM_FENCE); - - __local u32 s_msg_buf[128]; - - const u32 lid2 = lid * 2; - - s_msg_buf[lid2 + 0] = swap32 (ikepsk_bufs[salt_pos].msg_buf[lid2 + 0]); - s_msg_buf[lid2 + 1] = swap32 (ikepsk_bufs[salt_pos].msg_buf[lid2 + 1]); - - barrier (CLK_LOCAL_MEM_FENCE); - - if (gid >= gid_max) return; - /** * main */ - m05400s (w_s, 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, ikepsk_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, s_msg_buf); + m05400s (w_s, 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, ikepsk_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, s_msg_buf); }