X-Git-Url: https://www.flypig.org.uk/git/?a=blobdiff_plain;f=OpenCL%2Fm01460_a0.cl;h=64fcb1111212ca1eecfae4db6eb69542aee19c02;hb=161a6eb4bc643d8e636e96eda613f5137d30da59;hp=0abc9d929e1f8944031bbaf9c330cccda87ccd56;hpb=331188167cc7b4f6f2a336d63daca23fbdfc199c;p=hashcat.git

diff --git a/OpenCL/m01460_a0.cl b/OpenCL/m01460_a0.cl
index 0abc9d9..64fcb11 100644
--- a/OpenCL/m01460_a0.cl
+++ b/OpenCL/m01460_a0.cl
@@ -5,22 +5,16 @@
 
 #define _SHA256_
 
-#include "include/constants.h"
-#include "include/kernel_vendor.h"
+#define NEW_SIMD_CODE
 
-#define DGST_R0 3
-#define DGST_R1 7
-#define DGST_R2 2
-#define DGST_R3 6
-
-#include "include/kernel_functions.c"
-#include "OpenCL/types_ocl.c"
-#include "OpenCL/common.c"
-#include "include/rp_kernel.h"
-#include "OpenCL/rp.c"
-
-#define COMPARE_S "OpenCL/check_single_comp4.c"
-#define COMPARE_M "OpenCL/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_rp.h"
+#include "inc_rp.cl"
+#include "inc_simd.cl"
 
 __constant u32 k_sha256[64] =
 {
@@ -42,33 +36,33 @@ __constant u32 k_sha256[64] =
   SHA256C3c, SHA256C3d, SHA256C3e, SHA256C3f,
 };
 
-static void sha256_transform (const u32 w0[4], const u32 w1[4], const u32 w2[4], const u32 w3[4], u32 digest[8])
+void sha256_transform (const u32x w0[4], const u32x w1[4], const u32x w2[4], const u32x w3[4], u32x digest[8])
 {
-  u32 a = digest[0];
-  u32 b = digest[1];
-  u32 c = digest[2];
-  u32 d = digest[3];
-  u32 e = digest[4];
-  u32 f = digest[5];
-  u32 g = digest[6];
-  u32 h = digest[7];
-
-  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 f = digest[5];
+  u32x g = digest[6];
+  u32x h = digest[7];
+
+  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];
 
   #define ROUND_EXPAND()                            \
   {                                                 \
@@ -112,7 +106,9 @@ static void sha256_transform (const u32 w0[4], const u32 w1[4], const u32 w2[4],
 
   ROUND_STEP (0);
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (int i = 16; i < 64; i += 16)
   {
     ROUND_EXPAND (); ROUND_STEP (i);
@@ -128,7 +124,7 @@ static void sha256_transform (const u32 w0[4], const u32 w1[4], const u32 w2[4],
   digest[7] += h;
 }
 
-static void hmac_sha256_pad (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], u32 ipad[8], u32 opad[8])
+void hmac_sha256_pad (u32x w0[4], u32x w1[4], u32x w2[4], u32x w3[4], u32x ipad[8], u32x opad[8])
 {
   w0[0] = w0[0] ^ 0x36363636;
   w0[1] = w0[1] ^ 0x36363636;
@@ -187,7 +183,7 @@ static void hmac_sha256_pad (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], u32 ipa
   sha256_transform (w0, w1, w2, w3, opad);
 }
 
-static void hmac_sha256_run (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], u32 ipad[8], u32 opad[8], u32 digest[8])
+void hmac_sha256_run (u32x w0[4], u32x w1[4], u32x w2[4], u32x w3[4], u32x ipad[8], u32x opad[8], u32x digest[8])
 {
   digest[0] = ipad[0];
   digest[1] = ipad[1];
@@ -229,7 +225,7 @@ static void hmac_sha256_run (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], u32 ipa
   sha256_transform (w0, w1, w2, w3, digest);
 }
 
-__kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m01460_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_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 rules_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
+__kernel void m01460_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
@@ -246,18 +242,16 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m01460_m04 (__glo
   if (gid >= gid_max) return;
 
   u32 pw_buf0[4];
-
-  pw_buf0[0] = pws[gid].i[ 0];
-  pw_buf0[1] = pws[gid].i[ 1];
-  pw_buf0[2] = pws[gid].i[ 2];
-  pw_buf0[3] = pws[gid].i[ 3];
-
   u32 pw_buf1[4];
 
-  pw_buf1[0] = pws[gid].i[ 4];
-  pw_buf1[1] = pws[gid].i[ 5];
-  pw_buf1[2] = pws[gid].i[ 6];
-  pw_buf1[3] = pws[gid].i[ 7];
+  pw_buf0[0] = pws[gid].i[0];
+  pw_buf0[1] = pws[gid].i[1];
+  pw_buf0[2] = pws[gid].i[2];
+  pw_buf0[3] = pws[gid].i[3];
+  pw_buf1[0] = pws[gid].i[4];
+  pw_buf1[1] = pws[gid].i[5];
+  pw_buf1[2] = pws[gid].i[6];
+  pw_buf1[3] = pws[gid].i[7];
 
   const u32 pw_len = pws[gid].pw_len;
 
@@ -266,53 +260,57 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m01460_m04 (__glo
    */
 
   u32 salt_buf0[4];
-
-  salt_buf0[0] = salt_bufs[salt_pos].salt_buf[ 0];
-  salt_buf0[1] = salt_bufs[salt_pos].salt_buf[ 1];
-  salt_buf0[2] = salt_bufs[salt_pos].salt_buf[ 2];
-  salt_buf0[3] = salt_bufs[salt_pos].salt_buf[ 3];
-
   u32 salt_buf1[4];
-
-  salt_buf1[0] = salt_bufs[salt_pos].salt_buf[ 4];
-  salt_buf1[1] = salt_bufs[salt_pos].salt_buf[ 5];
-  salt_buf1[2] = salt_bufs[salt_pos].salt_buf[ 6];
-  salt_buf1[3] = salt_bufs[salt_pos].salt_buf[ 7];
+  u32 salt_buf2[4];
+  u32 salt_buf3[4];
+
+  salt_buf0[0] = swap32_S (salt_bufs[salt_pos].salt_buf[ 0]);
+  salt_buf0[1] = swap32_S (salt_bufs[salt_pos].salt_buf[ 1]);
+  salt_buf0[2] = swap32_S (salt_bufs[salt_pos].salt_buf[ 2]);
+  salt_buf0[3] = swap32_S (salt_bufs[salt_pos].salt_buf[ 3]);
+  salt_buf1[0] = swap32_S (salt_bufs[salt_pos].salt_buf[ 4]);
+  salt_buf1[1] = swap32_S (salt_bufs[salt_pos].salt_buf[ 5]);
+  salt_buf1[2] = swap32_S (salt_bufs[salt_pos].salt_buf[ 6]);
+  salt_buf1[3] = swap32_S (salt_bufs[salt_pos].salt_buf[ 7]);
+  salt_buf2[0] = swap32_S (salt_bufs[salt_pos].salt_buf[ 8]);
+  salt_buf2[1] = swap32_S (salt_bufs[salt_pos].salt_buf[ 9]);
+  salt_buf2[2] = swap32_S (salt_bufs[salt_pos].salt_buf[10]);
+  salt_buf2[3] = swap32_S (salt_bufs[salt_pos].salt_buf[11]);
+  salt_buf3[0] = swap32_S (salt_bufs[salt_pos].salt_buf[12]);
+  salt_buf3[1] = swap32_S (salt_bufs[salt_pos].salt_buf[13]);
+  salt_buf3[2] = swap32_S (salt_bufs[salt_pos].salt_buf[14]);
+  salt_buf3[3] = swap32_S (salt_bufs[salt_pos].salt_buf[15]);
+
+  const u32 salt_len = salt_bufs[salt_pos].salt_len;
 
   /**
    * pads
    */
 
-  u32 w0_t[4];
-
-  w0_t[0] = swap32 (salt_buf0[0]);
-  w0_t[1] = swap32 (salt_buf0[1]);
-  w0_t[2] = swap32 (salt_buf0[2]);
-  w0_t[3] = swap32 (salt_buf0[3]);
-
-  u32 w1_t[4];
-
-  w1_t[0] = swap32 (salt_buf1[0]);
-  w1_t[1] = swap32 (salt_buf1[1]);
-  w1_t[2] = swap32 (salt_buf1[2]);
-  w1_t[3] = swap32 (salt_buf1[3]);
-
-  u32 w2_t[4];
-
-  w2_t[0] = 0;
-  w2_t[1] = 0;
-  w2_t[2] = 0;
-  w2_t[3] = 0;
-
-  u32 w3_t[4];
-
-  w3_t[0] = 0;
-  w3_t[1] = 0;
-  w3_t[2] = 0;
-  w3_t[3] = 0;
-
-  u32 ipad[8];
-  u32 opad[8];
+  u32x w0_t[4];
+  u32x w1_t[4];
+  u32x w2_t[4];
+  u32x w3_t[4];
+
+  w0_t[0] = salt_buf0[0];
+  w0_t[1] = salt_buf0[1];
+  w0_t[2] = salt_buf0[2];
+  w0_t[3] = salt_buf0[3];
+  w1_t[0] = salt_buf1[0];
+  w1_t[1] = salt_buf1[1];
+  w1_t[2] = salt_buf1[2];
+  w1_t[3] = salt_buf1[3];
+  w2_t[0] = salt_buf2[0];
+  w2_t[1] = salt_buf2[1];
+  w2_t[2] = salt_buf2[2];
+  w2_t[3] = salt_buf2[3];
+  w3_t[0] = salt_buf3[0];
+  w3_t[1] = salt_buf3[1];
+  w3_t[2] = salt_buf3[2];
+  w3_t[3] = salt_buf3[3];
+
+  u32x ipad[8];
+  u32x opad[8];
 
   hmac_sha256_pad (w0_t, w1_t, w2_t, w3_t, ipad, opad);
 
@@ -320,39 +318,16 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m01460_m04 (__glo
    * loop
    */
 
-  for (u32 il_pos = 0; il_pos < rules_cnt; il_pos++)
+  for (u32 il_pos = 0; il_pos < il_cnt; il_pos += VECT_SIZE)
   {
-    u32 w0[4];
-
-    w0[0] = pw_buf0[0];
-    w0[1] = pw_buf0[1];
-    w0[2] = pw_buf0[2];
-    w0[3] = pw_buf0[3];
-
-    u32 w1[4];
-
-    w1[0] = pw_buf1[0];
-    w1[1] = pw_buf1[1];
-    w1[2] = pw_buf1[2];
-    w1[3] = pw_buf1[3];
-
-    u32 w2[4];
-
-    w2[0] = 0;
-    w2[1] = 0;
-    w2[2] = 0;
-    w2[3] = 0;
-
-    u32 w3[4];
-
-    w3[0] = 0;
-    w3[1] = 0;
-    w3[2] = 0;
-    w3[3] = 0;
+    u32x w0[4] = { 0 };
+    u32x w1[4] = { 0 };
+    u32x w2[4] = { 0 };
+    u32x w3[4] = { 0 };
 
-    const u32 out_len = apply_rules (rules_buf[il_pos].cmds, w0, w1, pw_len);
+    const u32x out_len = apply_rules_vect (pw_buf0, pw_buf1, pw_len, rules_buf, il_pos, w0, w1);
 
-    append_0x80_2x4 (w0, w1, out_len);
+    append_0x80_2x4_VV (w0, w1, out_len);
 
     w0_t[0] = swap32 (w0[0]);
     w0_t[1] = swap32 (w0[1]);
@@ -371,28 +346,23 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m01460_m04 (__glo
     w3_t[2] = 0;
     w3_t[3] = (64 + out_len) * 8;
 
-    u32 digest[8];
+    u32x digest[8];
 
     hmac_sha256_run (w0_t, w1_t, w2_t, w3_t, ipad, opad, digest);
 
-    const u32 r0 = digest[3];
-    const u32 r1 = digest[7];
-    const u32 r2 = digest[2];
-    const u32 r3 = digest[6];
-
-    #include COMPARE_M
+    COMPARE_M_SIMD (digest[3], digest[7], digest[2], digest[6]);
   }
 }
 
-__kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m01460_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_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 combs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
+__kernel void m01460_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)
 {
 }
 
-__kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m01460_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_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 combs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
+__kernel void m01460_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)
 {
 }
 
-__kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m01460_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_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 rules_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
+__kernel void m01460_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
@@ -409,18 +379,16 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m01460_s04 (__glo
   if (gid >= gid_max) return;
 
   u32 pw_buf0[4];
-
-  pw_buf0[0] = pws[gid].i[ 0];
-  pw_buf0[1] = pws[gid].i[ 1];
-  pw_buf0[2] = pws[gid].i[ 2];
-  pw_buf0[3] = pws[gid].i[ 3];
-
   u32 pw_buf1[4];
 
-  pw_buf1[0] = pws[gid].i[ 4];
-  pw_buf1[1] = pws[gid].i[ 5];
-  pw_buf1[2] = pws[gid].i[ 6];
-  pw_buf1[3] = pws[gid].i[ 7];
+  pw_buf0[0] = pws[gid].i[0];
+  pw_buf0[1] = pws[gid].i[1];
+  pw_buf0[2] = pws[gid].i[2];
+  pw_buf0[3] = pws[gid].i[3];
+  pw_buf1[0] = pws[gid].i[4];
+  pw_buf1[1] = pws[gid].i[5];
+  pw_buf1[2] = pws[gid].i[6];
+  pw_buf1[3] = pws[gid].i[7];
 
   const u32 pw_len = pws[gid].pw_len;
 
@@ -429,53 +397,57 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m01460_s04 (__glo
    */
 
   u32 salt_buf0[4];
-
-  salt_buf0[0] = salt_bufs[salt_pos].salt_buf[ 0];
-  salt_buf0[1] = salt_bufs[salt_pos].salt_buf[ 1];
-  salt_buf0[2] = salt_bufs[salt_pos].salt_buf[ 2];
-  salt_buf0[3] = salt_bufs[salt_pos].salt_buf[ 3];
-
   u32 salt_buf1[4];
-
-  salt_buf1[0] = salt_bufs[salt_pos].salt_buf[ 4];
-  salt_buf1[1] = salt_bufs[salt_pos].salt_buf[ 5];
-  salt_buf1[2] = salt_bufs[salt_pos].salt_buf[ 6];
-  salt_buf1[3] = salt_bufs[salt_pos].salt_buf[ 7];
+  u32 salt_buf2[4];
+  u32 salt_buf3[4];
+
+  salt_buf0[0] = swap32_S (salt_bufs[salt_pos].salt_buf[ 0]);
+  salt_buf0[1] = swap32_S (salt_bufs[salt_pos].salt_buf[ 1]);
+  salt_buf0[2] = swap32_S (salt_bufs[salt_pos].salt_buf[ 2]);
+  salt_buf0[3] = swap32_S (salt_bufs[salt_pos].salt_buf[ 3]);
+  salt_buf1[0] = swap32_S (salt_bufs[salt_pos].salt_buf[ 4]);
+  salt_buf1[1] = swap32_S (salt_bufs[salt_pos].salt_buf[ 5]);
+  salt_buf1[2] = swap32_S (salt_bufs[salt_pos].salt_buf[ 6]);
+  salt_buf1[3] = swap32_S (salt_bufs[salt_pos].salt_buf[ 7]);
+  salt_buf2[0] = swap32_S (salt_bufs[salt_pos].salt_buf[ 8]);
+  salt_buf2[1] = swap32_S (salt_bufs[salt_pos].salt_buf[ 9]);
+  salt_buf2[2] = swap32_S (salt_bufs[salt_pos].salt_buf[10]);
+  salt_buf2[3] = swap32_S (salt_bufs[salt_pos].salt_buf[11]);
+  salt_buf3[0] = swap32_S (salt_bufs[salt_pos].salt_buf[12]);
+  salt_buf3[1] = swap32_S (salt_bufs[salt_pos].salt_buf[13]);
+  salt_buf3[2] = swap32_S (salt_bufs[salt_pos].salt_buf[14]);
+  salt_buf3[3] = swap32_S (salt_bufs[salt_pos].salt_buf[15]);
+
+  const u32 salt_len = salt_bufs[salt_pos].salt_len;
 
   /**
    * pads
    */
 
-  u32 w0_t[4];
-
-  w0_t[0] = swap32 (salt_buf0[0]);
-  w0_t[1] = swap32 (salt_buf0[1]);
-  w0_t[2] = swap32 (salt_buf0[2]);
-  w0_t[3] = swap32 (salt_buf0[3]);
-
-  u32 w1_t[4];
-
-  w1_t[0] = swap32 (salt_buf1[0]);
-  w1_t[1] = swap32 (salt_buf1[1]);
-  w1_t[2] = swap32 (salt_buf1[2]);
-  w1_t[3] = swap32 (salt_buf1[3]);
-
-  u32 w2_t[4];
-
-  w2_t[0] = 0;
-  w2_t[1] = 0;
-  w2_t[2] = 0;
-  w2_t[3] = 0;
-
-  u32 w3_t[4];
-
-  w3_t[0] = 0;
-  w3_t[1] = 0;
-  w3_t[2] = 0;
-  w3_t[3] = 0;
-
-  u32 ipad[8];
-  u32 opad[8];
+  u32x w0_t[4];
+  u32x w1_t[4];
+  u32x w2_t[4];
+  u32x w3_t[4];
+
+  w0_t[0] = salt_buf0[0];
+  w0_t[1] = salt_buf0[1];
+  w0_t[2] = salt_buf0[2];
+  w0_t[3] = salt_buf0[3];
+  w1_t[0] = salt_buf1[0];
+  w1_t[1] = salt_buf1[1];
+  w1_t[2] = salt_buf1[2];
+  w1_t[3] = salt_buf1[3];
+  w2_t[0] = salt_buf2[0];
+  w2_t[1] = salt_buf2[1];
+  w2_t[2] = salt_buf2[2];
+  w2_t[3] = salt_buf2[3];
+  w3_t[0] = salt_buf3[0];
+  w3_t[1] = salt_buf3[1];
+  w3_t[2] = salt_buf3[2];
+  w3_t[3] = salt_buf3[3];
+
+  u32x ipad[8];
+  u32x opad[8];
 
   hmac_sha256_pad (w0_t, w1_t, w2_t, w3_t, ipad, opad);
 
@@ -495,39 +467,16 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m01460_s04 (__glo
    * loop
    */
 
-  for (u32 il_pos = 0; il_pos < rules_cnt; il_pos++)
+  for (u32 il_pos = 0; il_pos < il_cnt; il_pos += VECT_SIZE)
   {
-    u32 w0[4];
-
-    w0[0] = pw_buf0[0];
-    w0[1] = pw_buf0[1];
-    w0[2] = pw_buf0[2];
-    w0[3] = pw_buf0[3];
-
-    u32 w1[4];
-
-    w1[0] = pw_buf1[0];
-    w1[1] = pw_buf1[1];
-    w1[2] = pw_buf1[2];
-    w1[3] = pw_buf1[3];
+    u32x w0[4] = { 0 };
+    u32x w1[4] = { 0 };
+    u32x w2[4] = { 0 };
+    u32x w3[4] = { 0 };
 
-    u32 w2[4];
+    const u32x out_len = apply_rules_vect (pw_buf0, pw_buf1, pw_len, rules_buf, il_pos, w0, w1);
 
-    w2[0] = 0;
-    w2[1] = 0;
-    w2[2] = 0;
-    w2[3] = 0;
-
-    u32 w3[4];
-
-    w3[0] = 0;
-    w3[1] = 0;
-    w3[2] = 0;
-    w3[3] = 0;
-
-    const u32 out_len = apply_rules (rules_buf[il_pos].cmds, w0, w1, pw_len);
-
-    append_0x80_2x4 (w0, w1, out_len);
+    append_0x80_2x4_VV (w0, w1, out_len);
 
     w0_t[0] = swap32 (w0[0]);
     w0_t[1] = swap32 (w0[1]);
@@ -546,23 +495,18 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m01460_s04 (__glo
     w3_t[2] = 0;
     w3_t[3] = (64 + out_len) * 8;
 
-    u32 digest[8];
+    u32x digest[8];
 
     hmac_sha256_run (w0_t, w1_t, w2_t, w3_t, ipad, opad, digest);
 
-    const u32 r0 = digest[3];
-    const u32 r1 = digest[7];
-    const u32 r2 = digest[2];
-    const u32 r3 = digest[6];
-
-    #include COMPARE_S
+    COMPARE_S_SIMD (digest[3], digest[7], digest[2], digest[6]);
   }
 }
 
-__kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m01460_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_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 combs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
+__kernel void m01460_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)
 {
 }
 
-__kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m01460_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_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 combs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
+__kernel void m01460_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)
 {
 }