X-Git-Url: https://www.flypig.org.uk/git/?a=blobdiff_plain;f=OpenCL%2Fm09820_a0.cl;h=71c268ebbe16e242c3a87a4b54ef5d25a8a2724e;hb=161a6eb4bc643d8e636e96eda613f5137d30da59;hp=d6e3f12ed47ed7537bb71d6cc9d61213e5355ec2;hpb=0bf4e3c34a6a799ccc34f403bed70119574ca9c8;p=hashcat.git

diff --git a/OpenCL/m09820_a0.cl b/OpenCL/m09820_a0.cl
index d6e3f12..71c268e 100644
--- a/OpenCL/m09820_a0.cl
+++ b/OpenCL/m09820_a0.cl
@@ -5,47 +5,41 @@
 
 #define _OLDOFFICE34_
 
-#include "include/constants.h"
-#include "include/kernel_vendor.h"
-
-#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"
-#include "include/rp_gpu.h"
-#include "rp.c"
-
-#define COMPARE_S "check_single_comp4.c"
-#define COMPARE_M "check_multi_comp4.c"
-
-static void sha1_transform (const u32 w0[4], const u32 w1[4], const u32 w2[4], const u32 w3[4], u32 digest[5])
+#define NEW_SIMD_CODE
+
+#include "inc_vendor.cl"
+#include "inc_hash_constants.h"
+#include "inc_hash_functions.cl"
+#include "inc_types.cl"
+#include "inc_common.cl"
+#include "inc_rp.h"
+#include "inc_rp.cl"
+#include "inc_simd.cl"
+
+void sha1_transform (const u32x w0[4], const u32x w1[4], const u32x w2[4], const u32x w3[4], u32x digest[5])
 {
-  u32 A = digest[0];
-  u32 B = digest[1];
-  u32 C = digest[2];
-  u32 D = digest[3];
-  u32 E = digest[4];
-
-  u32 w0_t = w0[0];
-  u32 w1_t = w0[1];
-  u32 w2_t = w0[2];
-  u32 w3_t = w0[3];
-  u32 w4_t = w1[0];
-  u32 w5_t = w1[1];
-  u32 w6_t = w1[2];
-  u32 w7_t = w1[3];
-  u32 w8_t = w2[0];
-  u32 w9_t = w2[1];
-  u32 wa_t = w2[2];
-  u32 wb_t = w2[3];
-  u32 wc_t = w3[0];
-  u32 wd_t = w3[1];
-  u32 we_t = w3[2];
-  u32 wf_t = w3[3];
+  u32x A = digest[0];
+  u32x B = digest[1];
+  u32x C = digest[2];
+  u32x D = digest[3];
+  u32x E = digest[4];
+
+  u32x w0_t = w0[0];
+  u32x w1_t = w0[1];
+  u32x w2_t = w0[2];
+  u32x w3_t = w0[3];
+  u32x w4_t = w1[0];
+  u32x w5_t = w1[1];
+  u32x w6_t = w1[2];
+  u32x w7_t = w1[3];
+  u32x w8_t = w2[0];
+  u32x w9_t = w2[1];
+  u32x wa_t = w2[2];
+  u32x wb_t = w2[3];
+  u32x wc_t = w3[0];
+  u32x wd_t = w3[1];
+  u32x we_t = w3[2];
+  u32x wf_t = w3[3];
 
   #undef K
   #define K SHA1C00
@@ -150,7 +144,7 @@ static void sha1_transform (const u32 w0[4], const u32 w1[4], const u32 w2[4], c
   digest[4] += E;
 }
 
-__kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m09820_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 oldoffice34_t *oldoffice34_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 m09820_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 oldoffice34_t *oldoffice34_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
@@ -167,14 +161,12 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m09820_m04 (__glo
   if (gid >= gid_max) return;
 
   u32 pw_buf0[4];
+  u32 pw_buf1[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];
@@ -193,76 +185,46 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m09820_m04 (__glo
   salt_buf[2] = salt_bufs[salt_pos].salt_buf[2];
   salt_buf[3] = salt_bufs[salt_pos].salt_buf[3];
 
-  const u32 salt_len = 16;
-
-  const u32 pw_salt_len = pw_len + salt_len;
-
   /**
    * 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];
+    u32x w0[4] = { 0 };
+    u32x w1[4] = { 0 };
+    u32x w2[4] = { 0 };
+    u32x w3[4] = { 0 };
 
-    w0[0] = pw_buf0[0];
-    w0[1] = pw_buf0[1];
-    w0[2] = pw_buf0[2];
-    w0[3] = pw_buf0[3];
+    const u32x out_len = apply_rules_vect (pw_buf0, pw_buf1, pw_len, rules_buf, il_pos, w0, w1);
 
-    u32 w1[4];
+    append_0x80_2x4_VV (w0, w1, out_len);
 
-    w1[0] = pw_buf1[0];
-    w1[1] = pw_buf1[1];
-    w1[2] = pw_buf1[2];
-    w1[3] = pw_buf1[3];
+    /**
+     * sha1
+     */
 
-    u32 w2[4];
+    make_unicode (w1, w2, w3);
+    make_unicode (w0, w0, w1);
 
-    w2[0] = 0;
-    w2[1] = 0;
-    w2[2] = 0;
-    w2[3] = 0;
+    const u32x pw_salt_len = (out_len * 2) + 16;
 
-    u32 w3[4];
-
-    w3[0] = 0;
-    w3[1] = 0;
+    w3[3] = pw_salt_len * 8;
     w3[2] = 0;
-    w3[3] = 0;
-
-    const u32 out_len = apply_rules (rules_buf[il_pos].cmds, w0, w1, pw_len);
-
-    const u32 pw_salt_len = (out_len * 2) + salt_len;
-
-    append_0x80_2 (w0, w1, out_len);
-
-    u32 w0_t[4];
-    u32 w1_t[4];
-    u32 w2_t[4];
-    u32 w3_t[4];
-
-    make_unicode (w0, w0_t, w1_t);
-    make_unicode (w1, w2_t, w3_t);
-
-    switch_buffer_by_offset (w0_t, w1_t, w2_t, w3_t, salt_len);
-
-    w0_t[0] = salt_buf[0];
-    w0_t[1] = salt_buf[1];
-    w0_t[2] = salt_buf[2];
-    w0_t[3] = salt_buf[3];
-    w1_t[0] = swap_workaround (w1_t[0]);
-    w1_t[1] = swap_workaround (w1_t[1]);
-    w1_t[2] = swap_workaround (w1_t[2]);
-    w1_t[3] = swap_workaround (w1_t[3]);
-    w2_t[0] = swap_workaround (w2_t[0]);
-    w2_t[1] = swap_workaround (w2_t[1]);
-    w2_t[2] = swap_workaround (w2_t[2]);
-    w2_t[3] = swap_workaround (w2_t[3]);
-    w3_t[0] = swap_workaround (w3_t[0]);
-    w3_t[1] = swap_workaround (w3_t[1]);
-    w3_t[2] = 0;
-    w3_t[3] = pw_salt_len * 8;
+    w3[1] = swap32 (w2[1]);
+    w3[0] = swap32 (w2[0]);
+    w2[3] = swap32 (w1[3]);
+    w2[2] = swap32 (w1[2]);
+    w2[1] = swap32 (w1[1]);
+    w2[0] = swap32 (w1[0]);
+    w1[3] = swap32 (w0[3]);
+    w1[2] = swap32 (w0[2]);
+    w1[1] = swap32 (w0[1]);
+    w1[0] = swap32 (w0[0]);
+    w0[3] = salt_buf[3];
+    w0[2] = salt_buf[2];
+    w0[1] = salt_buf[1];
+    w0[0] = salt_buf[0];
 
     u32 digest[5];
 
@@ -272,24 +234,24 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m09820_m04 (__glo
     digest[3] = SHA1M_D;
     digest[4] = SHA1M_E;
 
-    sha1_transform (w0_t, w1_t, w2_t, w3_t, digest);
-
-    w0_t[0] = digest[0];
-    w0_t[1] = digest[1];
-    w0_t[2] = digest[2];
-    w0_t[3] = digest[3];
-    w1_t[0] = digest[4];
-    w1_t[1] = 0;
-    w1_t[2] = 0x80000000;
-    w1_t[3] = 0;
-    w2_t[0] = 0;
-    w2_t[1] = 0;
-    w2_t[2] = 0;
-    w2_t[3] = 0;
-    w3_t[0] = 0;
-    w3_t[1] = 0;
-    w3_t[2] = 0;
-    w3_t[3] = (20 + 4) * 8;
+    sha1_transform (w0, w1, w2, w3, digest);
+
+    w0[0] = digest[0];
+    w0[1] = digest[1];
+    w0[2] = digest[2];
+    w0[3] = digest[3];
+    w1[0] = digest[4];
+    w1[1] = 0;
+    w1[2] = 0x80000000;
+    w1[3] = 0;
+    w2[0] = 0;
+    w2[1] = 0;
+    w2[2] = 0;
+    w2[3] = 0;
+    w3[0] = 0;
+    w3[1] = 0;
+    w3[2] = 0;
+    w3[3] = (20 + 4) * 8;
 
     digest[0] = SHA1M_A;
     digest[1] = SHA1M_B;
@@ -297,29 +259,26 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m09820_m04 (__glo
     digest[3] = SHA1M_D;
     digest[4] = SHA1M_E;
 
-    sha1_transform (w0_t, w1_t, w2_t, w3_t, digest);
-
-    u32 a = swap_workaround (digest[0]);
-    u32 b = swap_workaround (digest[1]) & 0xff;
+    sha1_transform (w0, w1, w2, w3, digest);
 
-    const u32 r0 = a;
-    const u32 r1 = b;
-    const u32 r2 = 0;
-    const u32 r3 = 0;
+    digest[0] = swap32 (digest[0]);
+    digest[1] = swap32 (digest[1]) & 0xff;
+    digest[2] = 0;
+    digest[3] = 0;
 
-    #include COMPARE_M
+    COMPARE_M_SIMD (digest[0], digest[1], digest[2], digest[3]);
   }
 }
 
-__kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m09820_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 oldoffice34_t *oldoffice34_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 m09820_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 oldoffice34_t *oldoffice34_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))) m09820_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 oldoffice34_t *oldoffice34_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 m09820_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 oldoffice34_t *oldoffice34_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))) m09820_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 oldoffice34_t *oldoffice34_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 m09820_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 oldoffice34_t *oldoffice34_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
@@ -336,14 +295,12 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m09820_s04 (__glo
   if (gid >= gid_max) return;
 
   u32 pw_buf0[4];
+  u32 pw_buf1[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];
@@ -351,18 +308,6 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m09820_s04 (__glo
 
   const u32 pw_len = pws[gid].pw_len;
 
-  /**
-   * digest
-   */
-
-  const u32 search[4] =
-  {
-    digests_buf[digests_offset].digest_buf[DGST_R0],
-    digests_buf[digests_offset].digest_buf[DGST_R1],
-    digests_buf[digests_offset].digest_buf[DGST_R2],
-    digests_buf[digests_offset].digest_buf[DGST_R3]
-  };
-
   /**
    * salt
    */
@@ -374,76 +319,58 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m09820_s04 (__glo
   salt_buf[2] = salt_bufs[salt_pos].salt_buf[2];
   salt_buf[3] = salt_bufs[salt_pos].salt_buf[3];
 
-  const u32 salt_len = 16;
+  /**
+   * digest
+   */
 
-  const u32 pw_salt_len = pw_len + salt_len;
+  const u32 search[4] =
+  {
+    digests_buf[digests_offset].digest_buf[DGST_R0],
+    digests_buf[digests_offset].digest_buf[DGST_R1],
+    digests_buf[digests_offset].digest_buf[DGST_R2],
+    digests_buf[digests_offset].digest_buf[DGST_R3]
+  };
 
   /**
    * 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];
+    u32x w0[4] = { 0 };
+    u32x w1[4] = { 0 };
+    u32x w2[4] = { 0 };
+    u32x w3[4] = { 0 };
 
-    w0[0] = pw_buf0[0];
-    w0[1] = pw_buf0[1];
-    w0[2] = pw_buf0[2];
-    w0[3] = pw_buf0[3];
+    const u32x out_len = apply_rules_vect (pw_buf0, pw_buf1, pw_len, rules_buf, il_pos, w0, w1);
 
-    u32 w1[4];
+    append_0x80_2x4_VV (w0, w1, out_len);
 
-    w1[0] = pw_buf1[0];
-    w1[1] = pw_buf1[1];
-    w1[2] = pw_buf1[2];
-    w1[3] = pw_buf1[3];
+    /**
+     * sha1
+     */
 
-    u32 w2[4];
+    make_unicode (w1, w2, w3);
+    make_unicode (w0, w0, w1);
 
-    w2[0] = 0;
-    w2[1] = 0;
-    w2[2] = 0;
-    w2[3] = 0;
-
-    u32 w3[4];
+    const u32x pw_salt_len = (out_len * 2) + 16;
 
-    w3[0] = 0;
-    w3[1] = 0;
+    w3[3] = pw_salt_len * 8;
     w3[2] = 0;
-    w3[3] = 0;
-
-    const u32 out_len = apply_rules (rules_buf[il_pos].cmds, w0, w1, pw_len);
-
-    const u32 pw_salt_len = (out_len * 2) + salt_len;
-
-    append_0x80_2 (w0, w1, out_len);
-
-    u32 w0_t[4];
-    u32 w1_t[4];
-    u32 w2_t[4];
-    u32 w3_t[4];
-
-    make_unicode (w0, w0_t, w1_t);
-    make_unicode (w1, w2_t, w3_t);
-
-    switch_buffer_by_offset (w0_t, w1_t, w2_t, w3_t, salt_len);
-
-    w0_t[0] = salt_buf[0];
-    w0_t[1] = salt_buf[1];
-    w0_t[2] = salt_buf[2];
-    w0_t[3] = salt_buf[3];
-    w1_t[0] = swap_workaround (w1_t[0]);
-    w1_t[1] = swap_workaround (w1_t[1]);
-    w1_t[2] = swap_workaround (w1_t[2]);
-    w1_t[3] = swap_workaround (w1_t[3]);
-    w2_t[0] = swap_workaround (w2_t[0]);
-    w2_t[1] = swap_workaround (w2_t[1]);
-    w2_t[2] = swap_workaround (w2_t[2]);
-    w2_t[3] = swap_workaround (w2_t[3]);
-    w3_t[0] = swap_workaround (w3_t[0]);
-    w3_t[1] = swap_workaround (w3_t[1]);
-    w3_t[2] = 0;
-    w3_t[3] = pw_salt_len * 8;
+    w3[1] = swap32 (w2[1]);
+    w3[0] = swap32 (w2[0]);
+    w2[3] = swap32 (w1[3]);
+    w2[2] = swap32 (w1[2]);
+    w2[1] = swap32 (w1[1]);
+    w2[0] = swap32 (w1[0]);
+    w1[3] = swap32 (w0[3]);
+    w1[2] = swap32 (w0[2]);
+    w1[1] = swap32 (w0[1]);
+    w1[0] = swap32 (w0[0]);
+    w0[3] = salt_buf[3];
+    w0[2] = salt_buf[2];
+    w0[1] = salt_buf[1];
+    w0[0] = salt_buf[0];
 
     u32 digest[5];
 
@@ -453,24 +380,24 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m09820_s04 (__glo
     digest[3] = SHA1M_D;
     digest[4] = SHA1M_E;
 
-    sha1_transform (w0_t, w1_t, w2_t, w3_t, digest);
-
-    w0_t[0] = digest[0];
-    w0_t[1] = digest[1];
-    w0_t[2] = digest[2];
-    w0_t[3] = digest[3];
-    w1_t[0] = digest[4];
-    w1_t[1] = 0;
-    w1_t[2] = 0x80000000;
-    w1_t[3] = 0;
-    w2_t[0] = 0;
-    w2_t[1] = 0;
-    w2_t[2] = 0;
-    w2_t[3] = 0;
-    w3_t[0] = 0;
-    w3_t[1] = 0;
-    w3_t[2] = 0;
-    w3_t[3] = (20 + 4) * 8;
+    sha1_transform (w0, w1, w2, w3, digest);
+
+    w0[0] = digest[0];
+    w0[1] = digest[1];
+    w0[2] = digest[2];
+    w0[3] = digest[3];
+    w1[0] = digest[4];
+    w1[1] = 0;
+    w1[2] = 0x80000000;
+    w1[3] = 0;
+    w2[0] = 0;
+    w2[1] = 0;
+    w2[2] = 0;
+    w2[3] = 0;
+    w3[0] = 0;
+    w3[1] = 0;
+    w3[2] = 0;
+    w3[3] = (20 + 4) * 8;
 
     digest[0] = SHA1M_A;
     digest[1] = SHA1M_B;
@@ -478,24 +405,21 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m09820_s04 (__glo
     digest[3] = SHA1M_D;
     digest[4] = SHA1M_E;
 
-    sha1_transform (w0_t, w1_t, w2_t, w3_t, digest);
-
-    u32 a = swap_workaround (digest[0]);
-    u32 b = swap_workaround (digest[1]) & 0xff;
+    sha1_transform (w0, w1, w2, w3, digest);
 
-    const u32 r0 = a;
-    const u32 r1 = b;
-    const u32 r2 = 0;
-    const u32 r3 = 0;
+    digest[0] = swap32 (digest[0]);
+    digest[1] = swap32 (digest[1]) & 0xff;
+    digest[2] = 0;
+    digest[3] = 0;
 
-    #include COMPARE_S
+    COMPARE_S_SIMD (digest[0], digest[1], digest[2], digest[3]);
   }
 }
 
-__kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m09820_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 oldoffice34_t *oldoffice34_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 m09820_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 oldoffice34_t *oldoffice34_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))) m09820_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 oldoffice34_t *oldoffice34_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 m09820_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 oldoffice34_t *oldoffice34_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)
 {
 }