X-Git-Url: https://www.flypig.org.uk/git/?a=blobdiff_plain;f=OpenCL%2Fm01740_a0.cl;h=6a9d6649ff08f352b9fe2385a25ad90a5956d8e7;hb=161a6eb4bc643d8e636e96eda613f5137d30da59;hp=15f9389117774443953c66231b752da57463a113;hpb=fbc3a3517c7c25002c06d6680095ff66f50886e8;p=hashcat.git

diff --git a/OpenCL/m01740_a0.cl b/OpenCL/m01740_a0.cl
index 15f9389..6a9d664 100644
--- a/OpenCL/m01740_a0.cl
+++ b/OpenCL/m01740_a0.cl
@@ -5,22 +5,16 @@
 
 #define _SHA512_
 
-#include "include/constants.h"
-#include "include/kernel_vendor.h"
+#define NEW_SIMD_CODE
 
-#define DGST_R0 14
-#define DGST_R1 15
-#define DGST_R2 6
-#define DGST_R3 7
-
-#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 u64 k_sha512[80] =
 {
@@ -46,33 +40,33 @@ __constant u64 k_sha512[80] =
   SHA512C4c, SHA512C4d, SHA512C4e, SHA512C4f,
 };
 
-static void sha512_transform (const u32 w0[4], const u32 w1[4], const u32 w2[4], const u32 w3[4], u64 digest[8])
+void sha512_transform (const u32x w0[4], const u32x w1[4], const u32x w2[4], const u32x w3[4], u64x digest[8])
 {
-  u64 w0_t = hl32_to_64 (w0[0], w0[1]);
-  u64 w1_t = hl32_to_64 (w0[2], w0[3]);
-  u64 w2_t = hl32_to_64 (w1[0], w1[1]);
-  u64 w3_t = hl32_to_64 (w1[2], w1[3]);
-  u64 w4_t = hl32_to_64 (w2[0], w2[1]);
-  u64 w5_t = hl32_to_64 (w2[2], w2[3]);
-  u64 w6_t = hl32_to_64 (w3[0], w3[1]);
-  u64 w7_t = 0;
-  u64 w8_t = 0;
-  u64 w9_t = 0;
-  u64 wa_t = 0;
-  u64 wb_t = 0;
-  u64 wc_t = 0;
-  u64 wd_t = 0;
-  u64 we_t = 0;
-  u64 wf_t = hl32_to_64 (w3[2], w3[3]);
-
-  u64 a = digest[0];
-  u64 b = digest[1];
-  u64 c = digest[2];
-  u64 d = digest[3];
-  u64 e = digest[4];
-  u64 f = digest[5];
-  u64 g = digest[6];
-  u64 h = digest[7];
+  u64x w0_t = hl32_to_64 (w0[0], w0[1]);
+  u64x w1_t = hl32_to_64 (w0[2], w0[3]);
+  u64x w2_t = hl32_to_64 (w1[0], w1[1]);
+  u64x w3_t = hl32_to_64 (w1[2], w1[3]);
+  u64x w4_t = hl32_to_64 (w2[0], w2[1]);
+  u64x w5_t = hl32_to_64 (w2[2], w2[3]);
+  u64x w6_t = hl32_to_64 (w3[0], w3[1]);
+  u64x w7_t = 0;
+  u64x w8_t = 0;
+  u64x w9_t = 0;
+  u64x wa_t = 0;
+  u64x wb_t = 0;
+  u64x wc_t = 0;
+  u64x wd_t = 0;
+  u64x we_t = 0;
+  u64x wf_t = hl32_to_64 (w3[2], w3[3]);
+
+  u64x a = digest[0];
+  u64x b = digest[1];
+  u64x c = digest[2];
+  u64x d = digest[3];
+  u64x e = digest[4];
+  u64x f = digest[5];
+  u64x g = digest[6];
+  u64x h = digest[7];
 
   #define ROUND_EXPAND()                            \
   {                                                 \
@@ -116,7 +110,9 @@ static void sha512_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 < 80; i += 16)
   {
     ROUND_EXPAND (); ROUND_STEP (i);
@@ -143,7 +139,7 @@ static void sha512_transform (const u32 w0[4], const u32 w1[4], const u32 w2[4],
   digest[7] = h;
 }
 
-__kernel void m01740_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 m01740_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
@@ -160,18 +156,16 @@ __kernel void m01740_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf,
   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;
 
@@ -180,18 +174,26 @@ __kernel void m01740_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf,
    */
 
   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] = 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];
+  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];
+  salt_buf2[0] = salt_bufs[salt_pos].salt_buf[ 8];
+  salt_buf2[1] = salt_bufs[salt_pos].salt_buf[ 9];
+  salt_buf2[2] = salt_bufs[salt_pos].salt_buf[10];
+  salt_buf2[3] = salt_bufs[salt_pos].salt_buf[11];
+  salt_buf3[0] = salt_bufs[salt_pos].salt_buf[12];
+  salt_buf3[1] = salt_bufs[salt_pos].salt_buf[13];
+  salt_buf3[2] = salt_bufs[salt_pos].salt_buf[14];
+  salt_buf3[3] = salt_bufs[salt_pos].salt_buf[15];
 
   const u32 salt_len = salt_bufs[salt_pos].salt_len;
 
@@ -199,81 +201,74 @@ __kernel void m01740_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf,
    * 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];
-    u32 w1[4];
-    u32 w2[4];
-    u32 w3[4];
-
-    w0[0] = pw_buf0[0];
-    w0[1] = pw_buf0[1];
-    w0[2] = pw_buf0[2];
-    w0[3] = pw_buf0[3];
-    w1[0] = pw_buf1[0];
-    w1[1] = pw_buf1[1];
-    w1[2] = pw_buf1[2];
-    w1[3] = pw_buf1[3];
-    w2[0] = 0;
-    w2[1] = 0;
-    w2[2] = 0;
-    w2[3] = 0;
-    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);
+    u32x w0[4] = { 0 };
+    u32x w1[4] = { 0 };
+    u32x w2[4] = { 0 };
+    u32x w3[4] = { 0 };
 
-    /**
-     * prepend salt
-     */
-
-    const u32 out_salt_len = (out_len * 2) + salt_len;
-
-    u32 w0_t[4];
-    u32 w1_t[4];
-    u32 w2_t[4];
-    u32 w3_t[4];
+    const u32x out_len = apply_rules_vect (pw_buf0, pw_buf1, pw_len, rules_buf, il_pos, w0, w1);
 
-    make_unicode (w0, w0_t, w1_t);
-    make_unicode (w1, w2_t, w3_t);
+    make_unicode (w1, w2, w3);
+    make_unicode (w0, w0, w1);
 
-    switch_buffer_by_offset_le (w0_t, w1_t, w2_t, w3_t, salt_len);
+    const u32x out_len2 = out_len * 2;
 
-    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];
+    /**
+     * prepend salt
+     */
 
-    append_0x80_4x4 (w0_t, w1_t, w2_t, w3_t, out_salt_len);
+    const u32x out_salt_len = out_len2 + salt_len;
+
+    switch_buffer_by_offset_le_VV (w0, w1, w2, w3, salt_len);
+
+    w0[0] |= salt_buf0[0];
+    w0[1] |= salt_buf0[1];
+    w0[2] |= salt_buf0[2];
+    w0[3] |= salt_buf0[3];
+    w1[0] |= salt_buf1[0];
+    w1[1] |= salt_buf1[1];
+    w1[2] |= salt_buf1[2];
+    w1[3] |= salt_buf1[3];
+    w2[0] |= salt_buf2[0];
+    w2[1] |= salt_buf2[1];
+    w2[2] |= salt_buf2[2];
+    w2[3] |= salt_buf2[3];
+    w3[0] |= salt_buf3[0];
+    w3[1] |= salt_buf3[1];
+    w3[2] |= salt_buf3[2];
+    w3[3] |= salt_buf3[3];
+
+    append_0x80_4x4_VV (w0, w1, w2, w3, out_salt_len);
 
     /**
      * sha512
      */
 
-    w0_t[0] = swap32 (w0_t[0]);
-    w0_t[1] = swap32 (w0_t[1]);
-    w0_t[2] = swap32 (w0_t[2]);
-    w0_t[3] = swap32 (w0_t[3]);
-    w1_t[0] = swap32 (w1_t[0]);
-    w1_t[1] = swap32 (w1_t[1]);
-    w1_t[2] = swap32 (w1_t[2]);
-    w1_t[3] = swap32 (w1_t[3]);
-    w2_t[0] = swap32 (w2_t[0]);
-    w2_t[1] = swap32 (w2_t[1]);
-    w2_t[2] = swap32 (w2_t[2]);
-    w2_t[3] = swap32 (w2_t[3]);
-    w3_t[0] = swap32 (w3_t[0]);
-    w3_t[1] = swap32 (w3_t[1]);
+    u32x w0_t[4];
+    u32x w1_t[4];
+    u32x w2_t[4];
+    u32x w3_t[4];
+
+    w0_t[0] = swap32 (w0[0]);
+    w0_t[1] = swap32 (w0[1]);
+    w0_t[2] = swap32 (w0[2]);
+    w0_t[3] = swap32 (w0[3]);
+    w1_t[0] = swap32 (w1[0]);
+    w1_t[1] = swap32 (w1[1]);
+    w1_t[2] = swap32 (w1[2]);
+    w1_t[3] = swap32 (w1[3]);
+    w2_t[0] = swap32 (w2[0]);
+    w2_t[1] = swap32 (w2[1]);
+    w2_t[2] = swap32 (w2[2]);
+    w2_t[3] = swap32 (w2[3]);
+    w3_t[0] = swap32 (w3[0]);
+    w3_t[1] = swap32 (w3[1]);
     w3_t[2] = 0;
     w3_t[3] = out_salt_len * 8;
 
-    u64 digest[8];
+    u64x digest[8];
 
     digest[0] = SHA512M_A;
     digest[1] = SHA512M_B;
@@ -286,25 +281,24 @@ __kernel void m01740_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf,
 
     sha512_transform (w0_t, w1_t, w2_t, w3_t, digest);
 
+    const u32x r0 = l32_from_64 (digest[7]);
+    const u32x r1 = h32_from_64 (digest[7]);
+    const u32x r2 = l32_from_64 (digest[3]);
+    const u32x r3 = h32_from_64 (digest[3]);
 
-    const u32 r0 = l32_from_64 (digest[7]);
-    const u32 r1 = h32_from_64 (digest[7]);
-    const u32 r2 = l32_from_64 (digest[3]);
-    const u32 r3 = h32_from_64 (digest[3]);
-
-    #include COMPARE_M
+    COMPARE_M_SIMD (r0, r1, r2, r3);
   }
 }
 
-__kernel void m01740_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 rules_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
+__kernel void m01740_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 m01740_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 rules_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
+__kernel void m01740_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 m01740_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 m01740_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
@@ -321,18 +315,16 @@ __kernel void m01740_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf,
   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;
 
@@ -341,18 +333,26 @@ __kernel void m01740_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf,
    */
 
   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] = 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];
+  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];
+  salt_buf2[0] = salt_bufs[salt_pos].salt_buf[ 8];
+  salt_buf2[1] = salt_bufs[salt_pos].salt_buf[ 9];
+  salt_buf2[2] = salt_bufs[salt_pos].salt_buf[10];
+  salt_buf2[3] = salt_bufs[salt_pos].salt_buf[11];
+  salt_buf3[0] = salt_bufs[salt_pos].salt_buf[12];
+  salt_buf3[1] = salt_bufs[salt_pos].salt_buf[13];
+  salt_buf3[2] = salt_bufs[salt_pos].salt_buf[14];
+  salt_buf3[3] = salt_bufs[salt_pos].salt_buf[15];
 
   const u32 salt_len = salt_bufs[salt_pos].salt_len;
 
@@ -372,81 +372,74 @@ __kernel void m01740_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf,
    * 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];
-    u32 w1[4];
-    u32 w2[4];
-    u32 w3[4];
-
-    w0[0] = pw_buf0[0];
-    w0[1] = pw_buf0[1];
-    w0[2] = pw_buf0[2];
-    w0[3] = pw_buf0[3];
-    w1[0] = pw_buf1[0];
-    w1[1] = pw_buf1[1];
-    w1[2] = pw_buf1[2];
-    w1[3] = pw_buf1[3];
-    w2[0] = 0;
-    w2[1] = 0;
-    w2[2] = 0;
-    w2[3] = 0;
-    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);
+    u32x w0[4] = { 0 };
+    u32x w1[4] = { 0 };
+    u32x w2[4] = { 0 };
+    u32x w3[4] = { 0 };
 
-    /**
-     * prepend salt
-     */
-
-    const u32 out_salt_len = (out_len * 2) + salt_len;
+    const u32x out_len = apply_rules_vect (pw_buf0, pw_buf1, pw_len, rules_buf, il_pos, w0, w1);
 
-    u32 w0_t[4];
-    u32 w1_t[4];
-    u32 w2_t[4];
-    u32 w3_t[4];
+    make_unicode (w1, w2, w3);
+    make_unicode (w0, w0, w1);
 
-    make_unicode (w0, w0_t, w1_t);
-    make_unicode (w1, w2_t, w3_t);
+    const u32x out_len2 = out_len * 2;
 
-    switch_buffer_by_offset_le (w0_t, w1_t, w2_t, w3_t, salt_len);
-
-    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];
+    /**
+     * prepend salt
+     */
 
-    append_0x80_4x4 (w0_t, w1_t, w2_t, w3_t, out_salt_len);
+    const u32x out_salt_len = out_len2 + salt_len;
+
+    switch_buffer_by_offset_le_VV (w0, w1, w2, w3, salt_len);
+
+    w0[0] |= salt_buf0[0];
+    w0[1] |= salt_buf0[1];
+    w0[2] |= salt_buf0[2];
+    w0[3] |= salt_buf0[3];
+    w1[0] |= salt_buf1[0];
+    w1[1] |= salt_buf1[1];
+    w1[2] |= salt_buf1[2];
+    w1[3] |= salt_buf1[3];
+    w2[0] |= salt_buf2[0];
+    w2[1] |= salt_buf2[1];
+    w2[2] |= salt_buf2[2];
+    w2[3] |= salt_buf2[3];
+    w3[0] |= salt_buf3[0];
+    w3[1] |= salt_buf3[1];
+    w3[2] |= salt_buf3[2];
+    w3[3] |= salt_buf3[3];
+
+    append_0x80_4x4_VV (w0, w1, w2, w3, out_salt_len);
 
     /**
      * sha512
      */
 
-    w0_t[0] = swap32 (w0_t[0]);
-    w0_t[1] = swap32 (w0_t[1]);
-    w0_t[2] = swap32 (w0_t[2]);
-    w0_t[3] = swap32 (w0_t[3]);
-    w1_t[0] = swap32 (w1_t[0]);
-    w1_t[1] = swap32 (w1_t[1]);
-    w1_t[2] = swap32 (w1_t[2]);
-    w1_t[3] = swap32 (w1_t[3]);
-    w2_t[0] = swap32 (w2_t[0]);
-    w2_t[1] = swap32 (w2_t[1]);
-    w2_t[2] = swap32 (w2_t[2]);
-    w2_t[3] = swap32 (w2_t[3]);
-    w3_t[0] = swap32 (w3_t[0]);
-    w3_t[1] = swap32 (w3_t[1]);
+    u32x w0_t[4];
+    u32x w1_t[4];
+    u32x w2_t[4];
+    u32x w3_t[4];
+
+    w0_t[0] = swap32 (w0[0]);
+    w0_t[1] = swap32 (w0[1]);
+    w0_t[2] = swap32 (w0[2]);
+    w0_t[3] = swap32 (w0[3]);
+    w1_t[0] = swap32 (w1[0]);
+    w1_t[1] = swap32 (w1[1]);
+    w1_t[2] = swap32 (w1[2]);
+    w1_t[3] = swap32 (w1[3]);
+    w2_t[0] = swap32 (w2[0]);
+    w2_t[1] = swap32 (w2[1]);
+    w2_t[2] = swap32 (w2[2]);
+    w2_t[3] = swap32 (w2[3]);
+    w3_t[0] = swap32 (w3[0]);
+    w3_t[1] = swap32 (w3[1]);
     w3_t[2] = 0;
     w3_t[3] = out_salt_len * 8;
 
-    u64 digest[8];
+    u64x digest[8];
 
     digest[0] = SHA512M_A;
     digest[1] = SHA512M_B;
@@ -459,20 +452,19 @@ __kernel void m01740_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf,
 
     sha512_transform (w0_t, w1_t, w2_t, w3_t, digest);
 
+    const u32x r0 = l32_from_64 (digest[7]);
+    const u32x r1 = h32_from_64 (digest[7]);
+    const u32x r2 = l32_from_64 (digest[3]);
+    const u32x r3 = h32_from_64 (digest[3]);
 
-    const u32 r0 = l32_from_64 (digest[7]);
-    const u32 r1 = h32_from_64 (digest[7]);
-    const u32 r2 = l32_from_64 (digest[3]);
-    const u32 r3 = h32_from_64 (digest[3]);
-
-    #include COMPARE_S
+    COMPARE_S_SIMD (r0, r1, r2, r3);
   }
 }
 
-__kernel void m01740_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 rules_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
+__kernel void m01740_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 m01740_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 rules_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
+__kernel void m01740_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)
 {
 }