X-Git-Url: https://www.flypig.org.uk/git/?a=blobdiff_plain;f=OpenCL%2Fm05000_a0.cl;h=44a3e62e17fb2c00feed68e9f2abb405ffc86ce1;hb=161a6eb4bc643d8e636e96eda613f5137d30da59;hp=08371b44bd4ffb7c89051df355c4b6ef75aa7eac;hpb=0bf4e3c34a6a799ccc34f403bed70119574ca9c8;p=hashcat.git

diff --git a/OpenCL/m05000_a0.cl b/OpenCL/m05000_a0.cl
index 08371b4..44a3e62 100644
--- a/OpenCL/m05000_a0.cl
+++ b/OpenCL/m05000_a0.cl
@@ -5,22 +5,16 @@
 
 #define _KECCAK_
 
-#include "include/constants.h"
-#include "include/kernel_vendor.h"
+#define NEW_SIMD_CODE
 
-#define DGST_R0 2
-#define DGST_R1 3
-#define DGST_R2 4
-#define DGST_R3 5
-
-#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"
+#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 keccakf_rndc[24] =
 {
@@ -34,18 +28,6 @@ __constant u64 keccakf_rndc[24] =
   0x8000000000008080, 0x0000000080000001, 0x8000000080008008
 };
 
-__constant u32 keccakf_rotc[24] =
-{
-   1,  3,  6, 10, 15, 21, 28, 36, 45, 55,  2, 14,
-  27, 41, 56,  8, 25, 43, 62, 18, 39, 61, 20, 44
-};
-
-__constant u32 keccakf_piln[24] =
-{
-  10,  7, 11, 17, 18,  3,  5, 16,  8, 21, 24,  4,
-  15, 23, 19, 13, 12,  2, 20, 14, 22,  9,  6,  1
-};
-
 #ifndef KECCAK_ROUNDS
 #define KECCAK_ROUNDS 24
 #endif
@@ -63,8 +45,8 @@ __constant u32 keccakf_piln[24] =
 
 #define Rho_Pi(s)               \
 {                               \
-  u32 j = keccakf_piln[s];     \
-  u32 k = keccakf_rotc[s];     \
+  u32 j = keccakf_piln[s];      \
+  u32 k = keccakf_rotc[s];      \
   bc0 = st[j];                  \
   st[j] = rotl64 (t, k);        \
   t = bc0;                      \
@@ -84,7 +66,7 @@ __constant u32 keccakf_piln[24] =
   st[4 + s] ^= ~bc0 & bc1;      \
 }
 
-__kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m05000_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 rules_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
+__kernel void m05000_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
@@ -101,21 +83,35 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m05000_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;
 
+  /**
+   * constants
+   */
+
+  const u8 keccakf_rotc[24] =
+  {
+     1,  3,  6, 10, 15, 21, 28, 36, 45, 55,  2, 14,
+    27, 41, 56,  8, 25, 43, 62, 18, 39, 61, 20, 44
+  };
+
+  const u8 keccakf_piln[24] =
+  {
+    10,  7, 11, 17, 18,  3,  5, 16,  8, 21, 24,  4,
+    15, 23, 19, 13, 12,  2, 20, 14, 22,  9,  6,  1
+  };
+
   /**
    * 0x80 keccak, very special
    */
@@ -130,56 +126,27 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m05000_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];
+    u32x w0[4] = { 0 };
+    u32x w1[4] = { 0 };
+    u32x w2[4] = { 0 };
+    u32x w3[4] = { 0 };
 
-    w1[0] = pw_buf1[0];
-    w1[1] = pw_buf1[1];
-    w1[2] = pw_buf1[2];
-    w1[3] = pw_buf1[3];
+    const u32x out_len = apply_rules_vect (pw_buf0, pw_buf1, pw_len, rules_buf, il_pos, w0, w1);
 
-    u32 w2[4];
+    append_0x01_2x4_VV (w0, w1, out_len);
 
-    w2[0] = 0;
-    w2[1] = 0;
-    w2[2] = 0;
-    w2[3] = 0;
+    /**
+     * Keccak
+     */
 
-    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_0x01_2 (w0, w1, out_len);
-
-    u64 st[25];
-
-    #ifdef VECT_SIZE1
-    st[ 0] = (u64x) (w0[0])               | (u64x) (w0[1]) << 32;
-    st[ 1] = (u64x) (w0[2])               | (u64x) (w0[3]) << 32;
-    st[ 2] = (u64x) (w1[0])               | (u64x) (w1[1]) << 32;
-    st[ 3] = (u64x) (w1[2])               | (u64x) (w1[3]) << 32;
-    #endif
-
-    #ifdef VECT_SIZE2
-    st[ 0] = (u64x) (w0[0].s0, w0[0].s1)  | (u64x) (w0[1].s0, w0[1].s1) << 32;
-    st[ 1] = (u64x) (w0[2].s0, w0[2].s1)  | (u64x) (w0[3].s0, w0[3].s1) << 32;
-    st[ 2] = (u64x) (w1[0].s0, w1[0].s1)  | (u64x) (w1[1].s0, w1[1].s1) << 32;
-    st[ 3] = (u64x) (w1[2].s0, w1[2].s1)  | (u64x) (w1[3].s0, w1[3].s1) << 32;
-    #endif
+    u64x st[25];
 
+    st[ 0] = hl32_to_64 (w0[1], w0[0]);
+    st[ 1] = hl32_to_64 (w0[3], w0[2]);
+    st[ 2] = hl32_to_64 (w1[1], w1[0]);
+    st[ 3] = hl32_to_64 (w1[3], w1[2]);
     st[ 4] = 0;
     st[ 5] = 0;
     st[ 6] = 0;
@@ -210,13 +177,13 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m05000_m04 (__glo
     {
       // Theta
 
-      u64 bc0 = Theta1 (0);
-      u64 bc1 = Theta1 (1);
-      u64 bc2 = Theta1 (2);
-      u64 bc3 = Theta1 (3);
-      u64 bc4 = Theta1 (4);
+      u64x bc0 = Theta1 (0);
+      u64x bc1 = Theta1 (1);
+      u64x bc2 = Theta1 (2);
+      u64x bc3 = Theta1 (3);
+      u64x bc4 = Theta1 (4);
 
-      u64 t;
+      u64x t;
 
       t = bc4 ^ rotl64 (bc1, 1); Theta2 (0);
       t = bc0 ^ rotl64 (bc2, 1); Theta2 (1);
@@ -266,24 +233,24 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m05000_m04 (__glo
       st[0] ^= keccakf_rndc[round];
     }
 
-    const u32 r0 = l32_from_64 (st[1]);
-    const u32 r1 = h32_from_64 (st[1]);
-    const u32 r2 = l32_from_64 (st[2]);
-    const u32 r3 = h32_from_64 (st[2]);
+    const u32x r0 = l32_from_64 (st[1]);
+    const u32x r1 = h32_from_64 (st[1]);
+    const u32x r2 = l32_from_64 (st[2]);
+    const u32x r3 = h32_from_64 (st[2]);
 
-    #include COMPARE_M
+    COMPARE_M_SIMD (r0, r1, r2, r3);
   }
 }
 
-__kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m05000_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 combs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
+__kernel void m05000_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))) m05000_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 combs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
+__kernel void m05000_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))) m05000_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 rules_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
+__kernel void m05000_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
@@ -300,31 +267,33 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m05000_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;
 
   /**
-   * digest
+   * constants
    */
 
-  const u32 search[4] =
+  const u8 keccakf_rotc[24] =
   {
-    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]
+     1,  3,  6, 10, 15, 21, 28, 36, 45, 55,  2, 14,
+    27, 41, 56,  8, 25, 43, 62, 18, 39, 61, 20, 44
+  };
+
+  const u8 keccakf_piln[24] =
+  {
+    10,  7, 11, 17, 18,  3,  5, 16,  8, 21, 24,  4,
+    15, 23, 19, 13, 12,  2, 20, 14, 22,  9,  6,  1
   };
 
   /**
@@ -338,59 +307,42 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m05000_s04 (__glo
   const u32 add80w = (rsiz - 1) / 8;
 
   /**
-   * loop
+   * digest
    */
 
-  for (u32 il_pos = 0; il_pos < rules_cnt; il_pos++)
+  const u32 search[4] =
   {
-    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];
+    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]
+  };
 
-    w3[0] = 0;
-    w3[1] = 0;
-    w3[2] = 0;
-    w3[3] = 0;
+  /**
+   * loop
+   */
 
-    const u32 out_len = apply_rules (rules_buf[il_pos].cmds, w0, w1, pw_len);
+  for (u32 il_pos = 0; il_pos < il_cnt; il_pos += VECT_SIZE)
+  {
+    u32x w0[4] = { 0 };
+    u32x w1[4] = { 0 };
+    u32x w2[4] = { 0 };
+    u32x w3[4] = { 0 };
 
-    append_0x01_2 (w0, w1, out_len);
+    const u32x out_len = apply_rules_vect (pw_buf0, pw_buf1, pw_len, rules_buf, il_pos, w0, w1);
 
-    u64 st[25];
+    append_0x01_2x4_VV (w0, w1, out_len);
 
-    #ifdef VECT_SIZE1
-    st[ 0] = (u64x) (w0[0])               | (u64x) (w0[1]) << 32;
-    st[ 1] = (u64x) (w0[2])               | (u64x) (w0[3]) << 32;
-    st[ 2] = (u64x) (w1[0])               | (u64x) (w1[1]) << 32;
-    st[ 3] = (u64x) (w1[2])               | (u64x) (w1[3]) << 32;
-    #endif
+    /**
+     * Keccak
+     */
 
-    #ifdef VECT_SIZE2
-    st[ 0] = (u64x) (w0[0].s0, w0[0].s1)  | (u64x) (w0[1].s0, w0[1].s1) << 32;
-    st[ 1] = (u64x) (w0[2].s0, w0[2].s1)  | (u64x) (w0[3].s0, w0[3].s1) << 32;
-    st[ 2] = (u64x) (w1[0].s0, w1[0].s1)  | (u64x) (w1[1].s0, w1[1].s1) << 32;
-    st[ 3] = (u64x) (w1[2].s0, w1[2].s1)  | (u64x) (w1[3].s0, w1[3].s1) << 32;
-    #endif
+    u64x st[25];
 
+    st[ 0] = hl32_to_64 (w0[1], w0[0]);
+    st[ 1] = hl32_to_64 (w0[3], w0[2]);
+    st[ 2] = hl32_to_64 (w1[1], w1[0]);
+    st[ 3] = hl32_to_64 (w1[3], w1[2]);
     st[ 4] = 0;
     st[ 5] = 0;
     st[ 6] = 0;
@@ -421,13 +373,13 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m05000_s04 (__glo
     {
       // Theta
 
-      u64 bc0 = Theta1 (0);
-      u64 bc1 = Theta1 (1);
-      u64 bc2 = Theta1 (2);
-      u64 bc3 = Theta1 (3);
-      u64 bc4 = Theta1 (4);
+      u64x bc0 = Theta1 (0);
+      u64x bc1 = Theta1 (1);
+      u64x bc2 = Theta1 (2);
+      u64x bc3 = Theta1 (3);
+      u64x bc4 = Theta1 (4);
 
-      u64 t;
+      u64x t;
 
       t = bc4 ^ rotl64 (bc1, 1); Theta2 (0);
       t = bc0 ^ rotl64 (bc2, 1); Theta2 (1);
@@ -477,19 +429,19 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m05000_s04 (__glo
       st[0] ^= keccakf_rndc[round];
     }
 
-    const u32 r0 = l32_from_64 (st[1]);
-    const u32 r1 = h32_from_64 (st[1]);
-    const u32 r2 = l32_from_64 (st[2]);
-    const u32 r3 = h32_from_64 (st[2]);
+    const u32x r0 = l32_from_64 (st[1]);
+    const u32x r1 = h32_from_64 (st[1]);
+    const u32x r2 = l32_from_64 (st[2]);
+    const u32x r3 = h32_from_64 (st[2]);
 
-    #include COMPARE_S
+    COMPARE_S_SIMD (r0, r1, r2, r3);
   }
 }
 
-__kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m05000_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 combs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
+__kernel void m05000_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))) m05000_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 combs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
+__kernel void m05000_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)
 {
 }