X-Git-Url: https://www.flypig.org.uk/git/?a=blobdiff_plain;f=OpenCL%2Fm06100_a3.cl;h=e9f9d48f3e26c8fc70dd05c66c8d2fc747e0f9d3;hb=161a6eb4bc643d8e636e96eda613f5137d30da59;hp=1ae654d49fb76aef86abb0cbe7c9533f6d9bd45d;hpb=e4465d7dcb88426532f6b0488cf9bc1cd3b06cc6;p=hashcat.git

diff --git a/OpenCL/m06100_a3.cl b/OpenCL/m06100_a3.cl
index 1ae654d..e9f9d48 100644
--- a/OpenCL/m06100_a3.cl
+++ b/OpenCL/m06100_a3.cl
@@ -1,28 +1,35 @@
 /**
- * Author......: Jens Steube <jens.steube@gmail.com>
+ * Authors.....: Jens Steube <jens.steube@gmail.com>
+ *               Gabriele Gristina <matrix@hashcat.net>
+ *               magnum <john.magnum@hushmail.com>
+ *
  * License.....: MIT
  */
 
 #define _WHIRLPOOL_
 
-#include "include/constants.h"
-#include "include/kernel_vendor.h"
+#define NEW_SIMD_CODE
 
-#define DGST_R0 0
-#define DGST_R1 1
-#define DGST_R2 2
-#define DGST_R3 3
-
-#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"
+#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"
 
 #define R 10
 
+#if   VECT_SIZE == 1
 #define BOX(S,n,i) (S)[(n)][(i)]
+#elif VECT_SIZE == 2
+#define BOX(S,n,i) (u32x) ((S)[(n)][(i).s0], (S)[(n)][(i).s1])
+#elif VECT_SIZE == 4
+#define BOX(S,n,i) (u32x) ((S)[(n)][(i).s0], (S)[(n)][(i).s1], (S)[(n)][(i).s2], (S)[(n)][(i).s3])
+#elif VECT_SIZE == 8
+#define BOX(S,n,i) (u32x) ((S)[(n)][(i).s0], (S)[(n)][(i).s1], (S)[(n)][(i).s2], (S)[(n)][(i).s3], (S)[(n)][(i).s4], (S)[(n)][(i).s5], (S)[(n)][(i).s6], (S)[(n)][(i).s7])
+#elif VECT_SIZE == 16
+#define BOX(S,n,i) (u32x) ((S)[(n)][(i).s0], (S)[(n)][(i).s1], (S)[(n)][(i).s2], (S)[(n)][(i).s3], (S)[(n)][(i).s4], (S)[(n)][(i).s5], (S)[(n)][(i).s6], (S)[(n)][(i).s7], (S)[(n)][(i).s8], (S)[(n)][(i).s9], (S)[(n)][(i).sa], (S)[(n)][(i).sb], (S)[(n)][(i).sc], (S)[(n)][(i).sd], (S)[(n)][(i).se], (S)[(n)][(i).sf])
+#endif
 
 __constant u32 Ch[8][256] =
 {
@@ -1120,10 +1127,10 @@ __constant u32 rcl[R + 1] =
 
 // this is a highly optimized that assumes dgst[16] = { 0 }; only reuse of no 2nd transform is needed
 
-static void whirlpool_transform (const u32 w[16], u32 dgst[16], __local u32 s_Ch[8][256], __local u32 s_Cl[8][256])
+void whirlpool_transform (const u32x w[16], u32x dgst[16], __local u32 (*s_Ch)[256], __local u32 (*s_Cl)[256])
 {
-  u32 Kh[8];
-  u32 Kl[8];
+  u32x Kh[8];
+  u32x Kl[8];
 
   Kh[0] = 0x300beec0;
   Kl[0] = 0xaf902967;
@@ -1142,8 +1149,8 @@ static void whirlpool_transform (const u32 w[16], u32 dgst[16], __local u32 s_Ch
   Kh[7] = 0x28282828;
   Kl[7] = 0x28282828;
 
-  u32 stateh[8];
-  u32 statel[8];
+  u32x stateh[8];
+  u32x statel[8];
 
   stateh[0] = w[ 0];
   statel[0] = w[ 1];
@@ -1162,20 +1169,22 @@ static void whirlpool_transform (const u32 w[16], u32 dgst[16], __local u32 s_Ch
   stateh[7] = w[14];
   statel[7] = w[15];
 
-  u32 Lh[8];
-  u32 Ll[8];
+  u32x Lh[8];
+  u32x Ll[8];
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (int i = 0; i < 8; i++)
   {
-    const u32 Lp0 = stateh[(i + 8) & 7] >> 24;
-    const u32 Lp1 = stateh[(i + 7) & 7] >> 16;
-    const u32 Lp2 = stateh[(i + 6) & 7] >>  8;
-    const u32 Lp3 = stateh[(i + 5) & 7] >>  0;
-    const u32 Lp4 = statel[(i + 4) & 7] >> 24;
-    const u32 Lp5 = statel[(i + 3) & 7] >> 16;
-    const u32 Lp6 = statel[(i + 2) & 7] >>  8;
-    const u32 Lp7 = statel[(i + 1) & 7] >>  0;
+    const u32x Lp0 = stateh[(i + 8) & 7] >> 24;
+    const u32x Lp1 = stateh[(i + 7) & 7] >> 16;
+    const u32x Lp2 = stateh[(i + 6) & 7] >>  8;
+    const u32x Lp3 = stateh[(i + 5) & 7] >>  0;
+    const u32x Lp4 = statel[(i + 4) & 7] >> 24;
+    const u32x Lp5 = statel[(i + 3) & 7] >> 16;
+    const u32x Lp6 = statel[(i + 2) & 7] >>  8;
+    const u32x Lp7 = statel[(i + 1) & 7] >>  0;
 
     Lh[i] = BOX (s_Ch, 0, Lp0 & 0xff)
           ^ BOX (s_Ch, 1, Lp1 & 0xff)
@@ -1215,20 +1224,22 @@ static void whirlpool_transform (const u32 w[16], u32 dgst[16], __local u32 s_Ch
 
   for (int r = 2; r <= R; r++)
   {
-    u32 Lh[8];
-    u32 Ll[8];
+    u32x Lh[8];
+    u32x Ll[8];
 
+    #ifdef _unroll
     #pragma unroll
+    #endif
     for (int i = 0; i < 8; i++)
     {
-      const u32 Lp0 = Kh[(i + 8) & 7] >> 24;
-      const u32 Lp1 = Kh[(i + 7) & 7] >> 16;
-      const u32 Lp2 = Kh[(i + 6) & 7] >>  8;
-      const u32 Lp3 = Kh[(i + 5) & 7] >>  0;
-      const u32 Lp4 = Kl[(i + 4) & 7] >> 24;
-      const u32 Lp5 = Kl[(i + 3) & 7] >> 16;
-      const u32 Lp6 = Kl[(i + 2) & 7] >>  8;
-      const u32 Lp7 = Kl[(i + 1) & 7] >>  0;
+      const u32x Lp0 = Kh[(i + 8) & 7] >> 24;
+      const u32x Lp1 = Kh[(i + 7) & 7] >> 16;
+      const u32x Lp2 = Kh[(i + 6) & 7] >>  8;
+      const u32x Lp3 = Kh[(i + 5) & 7] >>  0;
+      const u32x Lp4 = Kl[(i + 4) & 7] >> 24;
+      const u32x Lp5 = Kl[(i + 3) & 7] >> 16;
+      const u32x Lp6 = Kl[(i + 2) & 7] >>  8;
+      const u32x Lp7 = Kl[(i + 1) & 7] >>  0;
 
       Lh[i] = BOX (s_Ch, 0, Lp0 & 0xff)
             ^ BOX (s_Ch, 1, Lp1 & 0xff)
@@ -1266,17 +1277,19 @@ static void whirlpool_transform (const u32 w[16], u32 dgst[16], __local u32 s_Ch
     Kh[7] = Lh[7];
     Kl[7] = Ll[7];
 
-    #pragma unroll 8
+    #ifdef _unroll
+    #pragma unroll
+    #endif
     for (int i = 0; i < 8; i++)
     {
-      const u32 Lp0 = stateh[(i + 8) & 7] >> 24;
-      const u32 Lp1 = stateh[(i + 7) & 7] >> 16;
-      const u32 Lp2 = stateh[(i + 6) & 7] >>  8;
-      const u32 Lp3 = stateh[(i + 5) & 7] >>  0;
-      const u32 Lp4 = statel[(i + 4) & 7] >> 24;
-      const u32 Lp5 = statel[(i + 3) & 7] >> 16;
-      const u32 Lp6 = statel[(i + 2) & 7] >>  8;
-      const u32 Lp7 = statel[(i + 1) & 7] >>  0;
+      const u32x Lp0 = stateh[(i + 8) & 7] >> 24;
+      const u32x Lp1 = stateh[(i + 7) & 7] >> 16;
+      const u32x Lp2 = stateh[(i + 6) & 7] >>  8;
+      const u32x Lp3 = stateh[(i + 5) & 7] >>  0;
+      const u32x Lp4 = statel[(i + 4) & 7] >> 24;
+      const u32x Lp5 = statel[(i + 3) & 7] >> 16;
+      const u32x Lp6 = statel[(i + 2) & 7] >>  8;
+      const u32x Lp7 = statel[(i + 1) & 7] >>  0;
 
       Lh[i] = BOX (s_Ch, 0, Lp0 & 0xff)
             ^ BOX (s_Ch, 1, Lp1 & 0xff)
@@ -1333,7 +1346,7 @@ static void whirlpool_transform (const u32 w[16], u32 dgst[16], __local u32 s_Ch
   dgst[15] = statel[7] ^ w[15];
 }
 
-static void m06100m (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 s_Cl[8][256], __local u32 s_Ch[8][256])
+void m06100m (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 (*s_Cl)[256], __local u32 (*s_Ch)[256])
 {
   /**
    * modifier
@@ -1348,45 +1361,61 @@ static void m06100m (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 wl[16];
-
-    wl[ 0] = w0[0];
-    wl[ 1] = w0[1];
-    wl[ 2] = w0[2];
-    wl[ 3] = w0[3];
-    wl[ 4] = w1[0];
-    wl[ 5] = w1[1];
-    wl[ 6] = w1[2];
-    wl[ 7] = w1[3];
-    wl[ 8] = w2[0];
-    wl[ 9] = w2[1];
-    wl[10] = w2[2];
-    wl[11] = w2[3];
-    wl[12] = w3[0];
-    wl[13] = w3[1];
-    wl[14] = 0;
-    wl[15] = pw_len * 8;
-
-    u32 dgst[16];
-
-    whirlpool_transform (wl, dgst, s_Ch, s_Cl);
-
-    const u32 r0 = dgst[0];
-    const u32 r1 = dgst[1];
-    const u32 r2 = dgst[2];
-    const u32 r3 = dgst[3];
-
-    #include COMPARE_M
+    const u32x w0r = ix_create_bft (bfs_buf, il_pos);
+
+    const u32x w0lr = w0l | w0r;
+
+    u32x w[16];
+
+    w[ 0] = w0lr;
+    w[ 1] = w0[1];
+    w[ 2] = w0[2];
+    w[ 3] = w0[3];
+    w[ 4] = w1[0];
+    w[ 5] = w1[1];
+    w[ 6] = w1[2];
+    w[ 7] = w1[3];
+    w[ 8] = w2[0];
+    w[ 9] = w2[1];
+    w[10] = w2[2];
+    w[11] = w2[3];
+    w[12] = w3[0];
+    w[13] = w3[1];
+    w[14] = 0;
+    w[15] = pw_len * 8;
+
+    /**
+     * Whirlool
+     */
+
+    u32x dgst[16];
+
+    dgst[ 0] = 0;
+    dgst[ 1] = 0;
+    dgst[ 2] = 0;
+    dgst[ 3] = 0;
+    dgst[ 4] = 0;
+    dgst[ 5] = 0;
+    dgst[ 6] = 0;
+    dgst[ 7] = 0;
+    dgst[ 8] = 0;
+    dgst[ 9] = 0;
+    dgst[10] = 0;
+    dgst[11] = 0;
+    dgst[12] = 0;
+    dgst[13] = 0;
+    dgst[14] = 0;
+    dgst[15] = 0;
+
+    whirlpool_transform (w, dgst, s_Ch, s_Cl);
+
+    COMPARE_M_SIMD (dgst[0], dgst[1], dgst[2], dgst[3]);
   }
 }
 
-static void m06100s (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 s_Cl[8][256], __local u32 s_Ch[8][256])
+void m06100s (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 (*s_Cl)[256], __local u32 (*s_Ch)[256])
 {
   /**
    * modifier
@@ -1413,57 +1442,105 @@ static void m06100s (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 wl[16];
-
-    wl[ 0] = w0[0];
-    wl[ 1] = w0[1];
-    wl[ 2] = w0[2];
-    wl[ 3] = w0[3];
-    wl[ 4] = w1[0];
-    wl[ 5] = w1[1];
-    wl[ 6] = w1[2];
-    wl[ 7] = w1[3];
-    wl[ 8] = w2[0];
-    wl[ 9] = w2[1];
-    wl[10] = w2[2];
-    wl[11] = w2[3];
-    wl[12] = w3[0];
-    wl[13] = w3[1];
-    wl[14] = 0;
-    wl[15] = pw_len * 8;
-
-    u32 dgst[16];
-
-    whirlpool_transform (wl, dgst, s_Ch, s_Cl);
-
-    const u32 r0 = dgst[0];
-    const u32 r1 = dgst[1];
-    const u32 r2 = dgst[2];
-    const u32 r3 = dgst[3];
-
-    #include COMPARE_S
+    const u32x w0r = ix_create_bft (bfs_buf, il_pos);
+
+    const u32x w0lr = w0l | w0r;
+
+    u32x w[16];
+
+    w[ 0] = w0lr;
+    w[ 1] = w0[1];
+    w[ 2] = w0[2];
+    w[ 3] = w0[3];
+    w[ 4] = w1[0];
+    w[ 5] = w1[1];
+    w[ 6] = w1[2];
+    w[ 7] = w1[3];
+    w[ 8] = w2[0];
+    w[ 9] = w2[1];
+    w[10] = w2[2];
+    w[11] = w2[3];
+    w[12] = w3[0];
+    w[13] = w3[1];
+    w[14] = 0;
+    w[15] = pw_len * 8;
+
+    /**
+     * Whirlool
+     */
+
+    u32x dgst[16];
+
+    dgst[ 0] = 0;
+    dgst[ 1] = 0;
+    dgst[ 2] = 0;
+    dgst[ 3] = 0;
+    dgst[ 4] = 0;
+    dgst[ 5] = 0;
+    dgst[ 6] = 0;
+    dgst[ 7] = 0;
+    dgst[ 8] = 0;
+    dgst[ 9] = 0;
+    dgst[10] = 0;
+    dgst[11] = 0;
+    dgst[12] = 0;
+    dgst[13] = 0;
+    dgst[14] = 0;
+    dgst[15] = 0;
+
+    whirlpool_transform (w, dgst, s_Ch, s_Cl);
+
+    COMPARE_S_SIMD (dgst[0], dgst[1], dgst[2], dgst[3]);
   }
 }
 
-__kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m06100_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 m06100_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)
 {
   /**
    * modifier
    */
 
+  const u32 gid = get_global_id (0);
   const u32 lid = get_local_id (0);
+  const u32 lsz = get_local_size (0);
 
   /**
-   * base
+   * shared
    */
 
-  const u32 gid = get_global_id (0);
+  __local u32 s_Ch[8][256];
+  __local u32 s_Cl[8][256];
+
+  for (u32 i = lid; i < 256; i += lsz)
+  {
+    s_Ch[0][i] = Ch[0][i];
+    s_Ch[1][i] = Ch[1][i];
+    s_Ch[2][i] = Ch[2][i];
+    s_Ch[3][i] = Ch[3][i];
+    s_Ch[4][i] = Ch[4][i];
+    s_Ch[5][i] = Ch[5][i];
+    s_Ch[6][i] = Ch[6][i];
+    s_Ch[7][i] = Ch[7][i];
+
+    s_Cl[0][i] = Cl[0][i];
+    s_Cl[1][i] = Cl[1][i];
+    s_Cl[2][i] = Cl[2][i];
+    s_Cl[3][i] = Cl[3][i];
+    s_Cl[4][i] = Cl[4][i];
+    s_Cl[5][i] = Cl[5][i];
+    s_Cl[6][i] = Cl[6][i];
+    s_Cl[7][i] = Cl[7][i];
+  }
+
+  barrier (CLK_LOCAL_MEM_FENCE);
+
+  if (gid >= gid_max) return;
+
+  /**
+   * base
+   */
 
   u32 w0[4];
 
@@ -1495,52 +1572,59 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m06100_m04 (__glo
 
   const u32 pw_len = pws[gid].pw_len;
 
+  /**
+   * main
+   */
+
+  m06100m (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, s_Cl, s_Ch);
+}
+
+__kernel void m06100_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)
+{
   /**
    * modifier
    */
 
+  const u32 gid = get_global_id (0);
+  const u32 lid = get_local_id (0);
+  const u32 lsz = get_local_size (0);
+
+  /**
+   * shared
+   */
+
   __local u32 s_Ch[8][256];
   __local u32 s_Cl[8][256];
 
-  const u32 lid4 = lid * 4;
-
-  for (u32 i = 0; i < 8; i++)
+  for (u32 i = lid; i < 256; i += lsz)
   {
-    s_Ch[i][lid4 + 0] = Ch[i][lid4 + 0];
-    s_Ch[i][lid4 + 1] = Ch[i][lid4 + 1];
-    s_Ch[i][lid4 + 2] = Ch[i][lid4 + 2];
-    s_Ch[i][lid4 + 3] = Ch[i][lid4 + 3];
-    s_Cl[i][lid4 + 0] = Cl[i][lid4 + 0];
-    s_Cl[i][lid4 + 1] = Cl[i][lid4 + 1];
-    s_Cl[i][lid4 + 2] = Cl[i][lid4 + 2];
-    s_Cl[i][lid4 + 3] = Cl[i][lid4 + 3];
+    s_Ch[0][i] = Ch[0][i];
+    s_Ch[1][i] = Ch[1][i];
+    s_Ch[2][i] = Ch[2][i];
+    s_Ch[3][i] = Ch[3][i];
+    s_Ch[4][i] = Ch[4][i];
+    s_Ch[5][i] = Ch[5][i];
+    s_Ch[6][i] = Ch[6][i];
+    s_Ch[7][i] = Ch[7][i];
+
+    s_Cl[0][i] = Cl[0][i];
+    s_Cl[1][i] = Cl[1][i];
+    s_Cl[2][i] = Cl[2][i];
+    s_Cl[3][i] = Cl[3][i];
+    s_Cl[4][i] = Cl[4][i];
+    s_Cl[5][i] = Cl[5][i];
+    s_Cl[6][i] = Cl[6][i];
+    s_Cl[7][i] = Cl[7][i];
   }
 
   barrier (CLK_LOCAL_MEM_FENCE);
 
   if (gid >= gid_max) return;
 
-  /**
-   * main
-   */
-
-  m06100m (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, s_Cl, s_Ch);
-}
-
-__kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m06100_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)
-{
-  /**
-   * 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];
@@ -1571,52 +1655,59 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m06100_m08 (__glo
 
   const u32 pw_len = pws[gid].pw_len;
 
+  /**
+   * main
+   */
+
+  m06100m (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, s_Cl, s_Ch);
+}
+
+__kernel void m06100_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)
+{
   /**
    * modifier
    */
 
+  const u32 gid = get_global_id (0);
+  const u32 lid = get_local_id (0);
+  const u32 lsz = get_local_size (0);
+
+  /**
+   * shared
+   */
+
   __local u32 s_Ch[8][256];
   __local u32 s_Cl[8][256];
 
-  const u32 lid4 = lid * 4;
-
-  for (u32 i = 0; i < 8; i++)
+  for (u32 i = lid; i < 256; i += lsz)
   {
-    s_Ch[i][lid4 + 0] = Ch[i][lid4 + 0];
-    s_Ch[i][lid4 + 1] = Ch[i][lid4 + 1];
-    s_Ch[i][lid4 + 2] = Ch[i][lid4 + 2];
-    s_Ch[i][lid4 + 3] = Ch[i][lid4 + 3];
-    s_Cl[i][lid4 + 0] = Cl[i][lid4 + 0];
-    s_Cl[i][lid4 + 1] = Cl[i][lid4 + 1];
-    s_Cl[i][lid4 + 2] = Cl[i][lid4 + 2];
-    s_Cl[i][lid4 + 3] = Cl[i][lid4 + 3];
+    s_Ch[0][i] = Ch[0][i];
+    s_Ch[1][i] = Ch[1][i];
+    s_Ch[2][i] = Ch[2][i];
+    s_Ch[3][i] = Ch[3][i];
+    s_Ch[4][i] = Ch[4][i];
+    s_Ch[5][i] = Ch[5][i];
+    s_Ch[6][i] = Ch[6][i];
+    s_Ch[7][i] = Ch[7][i];
+
+    s_Cl[0][i] = Cl[0][i];
+    s_Cl[1][i] = Cl[1][i];
+    s_Cl[2][i] = Cl[2][i];
+    s_Cl[3][i] = Cl[3][i];
+    s_Cl[4][i] = Cl[4][i];
+    s_Cl[5][i] = Cl[5][i];
+    s_Cl[6][i] = Cl[6][i];
+    s_Cl[7][i] = Cl[7][i];
   }
 
   barrier (CLK_LOCAL_MEM_FENCE);
 
   if (gid >= gid_max) return;
 
-  /**
-   * main
-   */
-
-  m06100m (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, s_Cl, s_Ch);
-}
-
-__kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m06100_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)
-{
-  /**
-   * 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];
@@ -1647,52 +1738,59 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m06100_m16 (__glo
 
   const u32 pw_len = pws[gid].pw_len;
 
+  /**
+   * main
+   */
+
+  m06100m (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, s_Cl, s_Ch);
+}
+
+__kernel void m06100_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)
+{
   /**
    * modifier
    */
 
+  const u32 gid = get_global_id (0);
+  const u32 lid = get_local_id (0);
+  const u32 lsz = get_local_size (0);
+
+  /**
+   * shared
+   */
+
   __local u32 s_Ch[8][256];
   __local u32 s_Cl[8][256];
 
-  const u32 lid4 = lid * 4;
-
-  for (u32 i = 0; i < 8; i++)
+  for (u32 i = lid; i < 256; i += lsz)
   {
-    s_Ch[i][lid4 + 0] = Ch[i][lid4 + 0];
-    s_Ch[i][lid4 + 1] = Ch[i][lid4 + 1];
-    s_Ch[i][lid4 + 2] = Ch[i][lid4 + 2];
-    s_Ch[i][lid4 + 3] = Ch[i][lid4 + 3];
-    s_Cl[i][lid4 + 0] = Cl[i][lid4 + 0];
-    s_Cl[i][lid4 + 1] = Cl[i][lid4 + 1];
-    s_Cl[i][lid4 + 2] = Cl[i][lid4 + 2];
-    s_Cl[i][lid4 + 3] = Cl[i][lid4 + 3];
+    s_Ch[0][i] = Ch[0][i];
+    s_Ch[1][i] = Ch[1][i];
+    s_Ch[2][i] = Ch[2][i];
+    s_Ch[3][i] = Ch[3][i];
+    s_Ch[4][i] = Ch[4][i];
+    s_Ch[5][i] = Ch[5][i];
+    s_Ch[6][i] = Ch[6][i];
+    s_Ch[7][i] = Ch[7][i];
+
+    s_Cl[0][i] = Cl[0][i];
+    s_Cl[1][i] = Cl[1][i];
+    s_Cl[2][i] = Cl[2][i];
+    s_Cl[3][i] = Cl[3][i];
+    s_Cl[4][i] = Cl[4][i];
+    s_Cl[5][i] = Cl[5][i];
+    s_Cl[6][i] = Cl[6][i];
+    s_Cl[7][i] = Cl[7][i];
   }
 
   barrier (CLK_LOCAL_MEM_FENCE);
 
   if (gid >= gid_max) return;
 
-  /**
-   * main
-   */
-
-  m06100m (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, s_Cl, s_Ch);
-}
-
-__kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m06100_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)
-{
-  /**
-   * 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];
@@ -1723,52 +1821,59 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m06100_s04 (__glo
 
   const u32 pw_len = pws[gid].pw_len;
 
+  /**
+   * main
+   */
+
+  m06100s (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, s_Cl, s_Ch);
+}
+
+__kernel void m06100_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)
+{
   /**
    * modifier
    */
 
+  const u32 gid = get_global_id (0);
+  const u32 lid = get_local_id (0);
+  const u32 lsz = get_local_size (0);
+
+  /**
+   * shared
+   */
+
   __local u32 s_Ch[8][256];
   __local u32 s_Cl[8][256];
 
-  const u32 lid4 = lid * 4;
-
-  for (u32 i = 0; i < 8; i++)
+  for (u32 i = lid; i < 256; i += lsz)
   {
-    s_Ch[i][lid4 + 0] = Ch[i][lid4 + 0];
-    s_Ch[i][lid4 + 1] = Ch[i][lid4 + 1];
-    s_Ch[i][lid4 + 2] = Ch[i][lid4 + 2];
-    s_Ch[i][lid4 + 3] = Ch[i][lid4 + 3];
-    s_Cl[i][lid4 + 0] = Cl[i][lid4 + 0];
-    s_Cl[i][lid4 + 1] = Cl[i][lid4 + 1];
-    s_Cl[i][lid4 + 2] = Cl[i][lid4 + 2];
-    s_Cl[i][lid4 + 3] = Cl[i][lid4 + 3];
+    s_Ch[0][i] = Ch[0][i];
+    s_Ch[1][i] = Ch[1][i];
+    s_Ch[2][i] = Ch[2][i];
+    s_Ch[3][i] = Ch[3][i];
+    s_Ch[4][i] = Ch[4][i];
+    s_Ch[5][i] = Ch[5][i];
+    s_Ch[6][i] = Ch[6][i];
+    s_Ch[7][i] = Ch[7][i];
+
+    s_Cl[0][i] = Cl[0][i];
+    s_Cl[1][i] = Cl[1][i];
+    s_Cl[2][i] = Cl[2][i];
+    s_Cl[3][i] = Cl[3][i];
+    s_Cl[4][i] = Cl[4][i];
+    s_Cl[5][i] = Cl[5][i];
+    s_Cl[6][i] = Cl[6][i];
+    s_Cl[7][i] = Cl[7][i];
   }
 
   barrier (CLK_LOCAL_MEM_FENCE);
 
   if (gid >= gid_max) return;
 
-  /**
-   * main
-   */
-
-  m06100s (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, s_Cl, s_Ch);
-}
-
-__kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m06100_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)
-{
-  /**
-   * 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];
@@ -1799,52 +1904,59 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m06100_s08 (__glo
 
   const u32 pw_len = pws[gid].pw_len;
 
+  /**
+   * main
+   */
+
+  m06100s (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, s_Cl, s_Ch);
+}
+
+__kernel void m06100_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)
+{
   /**
    * modifier
    */
 
+  const u32 gid = get_global_id (0);
+  const u32 lid = get_local_id (0);
+  const u32 lsz = get_local_size (0);
+
+  /**
+   * shared
+   */
+
   __local u32 s_Ch[8][256];
   __local u32 s_Cl[8][256];
 
-  const u32 lid4 = lid * 4;
-
-  for (u32 i = 0; i < 8; i++)
+  for (u32 i = lid; i < 256; i += lsz)
   {
-    s_Ch[i][lid4 + 0] = Ch[i][lid4 + 0];
-    s_Ch[i][lid4 + 1] = Ch[i][lid4 + 1];
-    s_Ch[i][lid4 + 2] = Ch[i][lid4 + 2];
-    s_Ch[i][lid4 + 3] = Ch[i][lid4 + 3];
-    s_Cl[i][lid4 + 0] = Cl[i][lid4 + 0];
-    s_Cl[i][lid4 + 1] = Cl[i][lid4 + 1];
-    s_Cl[i][lid4 + 2] = Cl[i][lid4 + 2];
-    s_Cl[i][lid4 + 3] = Cl[i][lid4 + 3];
+    s_Ch[0][i] = Ch[0][i];
+    s_Ch[1][i] = Ch[1][i];
+    s_Ch[2][i] = Ch[2][i];
+    s_Ch[3][i] = Ch[3][i];
+    s_Ch[4][i] = Ch[4][i];
+    s_Ch[5][i] = Ch[5][i];
+    s_Ch[6][i] = Ch[6][i];
+    s_Ch[7][i] = Ch[7][i];
+
+    s_Cl[0][i] = Cl[0][i];
+    s_Cl[1][i] = Cl[1][i];
+    s_Cl[2][i] = Cl[2][i];
+    s_Cl[3][i] = Cl[3][i];
+    s_Cl[4][i] = Cl[4][i];
+    s_Cl[5][i] = Cl[5][i];
+    s_Cl[6][i] = Cl[6][i];
+    s_Cl[7][i] = Cl[7][i];
   }
 
   barrier (CLK_LOCAL_MEM_FENCE);
 
   if (gid >= gid_max) return;
 
-  /**
-   * main
-   */
-
-  m06100s (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, s_Cl, s_Ch);
-}
-
-__kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m06100_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)
-{
-  /**
-   * 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];
@@ -1875,34 +1987,9 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m06100_s16 (__glo
 
   const u32 pw_len = pws[gid].pw_len;
 
-  /**
-   * modifier
-   */
-
-  __local u32 s_Ch[8][256];
-  __local u32 s_Cl[8][256];
-
-  const u32 lid4 = lid * 4;
-
-  for (u32 i = 0; i < 8; i++)
-  {
-    s_Ch[i][lid4 + 0] = Ch[i][lid4 + 0];
-    s_Ch[i][lid4 + 1] = Ch[i][lid4 + 1];
-    s_Ch[i][lid4 + 2] = Ch[i][lid4 + 2];
-    s_Ch[i][lid4 + 3] = Ch[i][lid4 + 3];
-    s_Cl[i][lid4 + 0] = Cl[i][lid4 + 0];
-    s_Cl[i][lid4 + 1] = Cl[i][lid4 + 1];
-    s_Cl[i][lid4 + 2] = Cl[i][lid4 + 2];
-    s_Cl[i][lid4 + 3] = Cl[i][lid4 + 3];
-  }
-
-  barrier (CLK_LOCAL_MEM_FENCE);
-
-  if (gid >= gid_max) return;
-
   /**
    * main
    */
 
-  m06100s (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, s_Cl, s_Ch);
+  m06100s (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, s_Cl, s_Ch);
 }