Remove some unroll to reduce register pressure

[hashcat.git] / OpenCL / m12300.cl

diff --git a/OpenCL/m12300.cl b/OpenCL/m12300.cl
index 97e67fe..5ab6a05 100644 (file)
--- a/OpenCL/m12300.cl
+++ b/OpenCL/m12300.cl
@@ -14,12 +14,11 @@
 #define DGST_R3 3
 
 #include "include/kernel_functions.c"
-#include "types_ocl.c"
-#include "common.c"
+#include "OpenCL/types_ocl.c"
+#include "OpenCL/common.c"
 
-#ifdef  VECT_SIZE1
-#define COMPARE_M "check_multi_vect1_comp4.c"
-#endif
+#define COMPARE_S "OpenCL/check_single_comp4.c"
+#define COMPARE_M "OpenCL/check_multi_comp4.c"
 
 __constant u64 k_sha512[80] =
 {
@@ -45,7 +44,7 @@ __constant u64 k_sha512[80] =
   SHA512C4c, SHA512C4d, SHA512C4e, SHA512C4f,
 };
 
-static void sha512_transform (const u64 w[16], u64 dgst[8])
+void sha512_transform (const u64 w[16], u64 dgst[8])
 {
   u64 a = dgst[0];
   u64 b = dgst[1];
@@ -115,7 +114,7 @@ static void sha512_transform (const u64 w[16], u64 dgst[8])
 
   ROUND_STEP (0);
 
-  //#pragma unroll // kernel fails if used
+  //#pragma unroll
   for (int i = 16; i < 80; i += 16)
   {
     ROUND_EXPAND (); ROUND_STEP (i);
@@ -131,7 +130,7 @@ static void sha512_transform (const u64 w[16], u64 dgst[8])
   dgst[7] += h;
 }
 
-static void hmac_run (const u64 w1[16], const u64 ipad[8], const u64 opad[8], u64 dgst[8])
+void hmac_run (const u64 w1[16], const u64 ipad[8], const u64 opad[8], u64 dgst[8])
 {
   dgst[0] = ipad[0];
   dgst[1] = ipad[1];
@@ -175,7 +174,7 @@ static void hmac_run (const u64 w1[16], const u64 ipad[8], const u64 opad[8], u6
   sha512_transform (w, dgst);
 }
 
-static void hmac_init (u64 w[16], u64 ipad[8], u64 opad[8])
+void hmac_init (u64 w[16], u64 ipad[8], u64 opad[8])
 {
   w[ 0] ^= 0x3636363636363636;
   w[ 1] ^= 0x3636363636363636;
@@ -234,7 +233,7 @@ static void hmac_init (u64 w[16], u64 ipad[8], u64 opad[8])
   sha512_transform (w, opad);
 }
 
-__kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m12300_init (__global pw_t *pws, __global gpu_rule_t *rules_buf, __global comb_t *combs_buf, __global bf_t *bfs_buf, __global oraclet_tmp_t *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 m12300_init (__global pw_t *pws, __global kernel_rule_t *rules_buf, __global comb_t *combs_buf, __global bf_t *bfs_buf, __global oraclet_tmp_t *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 il_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
 {
   /**
    * base
@@ -246,31 +245,31 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m12300_init (__gl
 
   u32 w0[4];
 
-  w0[0] = swap_workaround (pws[gid].i[ 0]);
-  w0[1] = swap_workaround (pws[gid].i[ 1]);
-  w0[2] = swap_workaround (pws[gid].i[ 2]);
-  w0[3] = swap_workaround (pws[gid].i[ 3]);
+  w0[0] = swap32 (pws[gid].i[ 0]);
+  w0[1] = swap32 (pws[gid].i[ 1]);
+  w0[2] = swap32 (pws[gid].i[ 2]);
+  w0[3] = swap32 (pws[gid].i[ 3]);
 
   u32 w1[4];
 
-  w1[0] = swap_workaround (pws[gid].i[ 4]);
-  w1[1] = swap_workaround (pws[gid].i[ 5]);
-  w1[2] = swap_workaround (pws[gid].i[ 6]);
-  w1[3] = swap_workaround (pws[gid].i[ 7]);
+  w1[0] = swap32 (pws[gid].i[ 4]);
+  w1[1] = swap32 (pws[gid].i[ 5]);
+  w1[2] = swap32 (pws[gid].i[ 6]);
+  w1[3] = swap32 (pws[gid].i[ 7]);
 
   u32 w2[4];
 
-  w2[0] = swap_workaround (pws[gid].i[ 8]);
-  w2[1] = swap_workaround (pws[gid].i[ 9]);
-  w2[2] = swap_workaround (pws[gid].i[10]);
-  w2[3] = swap_workaround (pws[gid].i[11]);
+  w2[0] = swap32 (pws[gid].i[ 8]);
+  w2[1] = swap32 (pws[gid].i[ 9]);
+  w2[2] = swap32 (pws[gid].i[10]);
+  w2[3] = swap32 (pws[gid].i[11]);
 
   u32 w3[4];
 
-  w3[0] = swap_workaround (pws[gid].i[12]);
-  w3[1] = swap_workaround (pws[gid].i[13]);
-  w3[2] = swap_workaround (pws[gid].i[14]);
-  w3[3] = swap_workaround (pws[gid].i[15]);
+  w3[0] = swap32 (pws[gid].i[12]);
+  w3[1] = swap32 (pws[gid].i[13]);
+  w3[2] = swap32 (pws[gid].i[14]);
+  w3[3] = swap32 (pws[gid].i[15]);
 
   /**
    * salt
@@ -363,7 +362,7 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m12300_init (__gl
   }
 }
 
-__kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m12300_loop (__global pw_t *pws, __global gpu_rule_t *rules_buf, __global comb_t *combs_buf, __global bf_t *bfs_buf, __global oraclet_tmp_t *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 m12300_loop (__global pw_t *pws, __global kernel_rule_t *rules_buf, __global comb_t *combs_buf, __global bf_t *bfs_buf, __global oraclet_tmp_t *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 il_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
 {
   const u32 gid = get_global_id (0);
 
@@ -468,7 +467,7 @@ __kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m12300_loop (__gl
   }
 }
 
-__kernel void __attribute__((reqd_work_group_size (64, 1, 1))) m12300_comp (__global pw_t *pws, __global gpu_rule_t *rules_buf, __global comb_t *combs_buf, __global bf_t *bfs_buf, __global oraclet_tmp_t *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 m12300_comp (__global pw_t *pws, __global kernel_rule_t *rules_buf, __global comb_t *combs_buf, __global bf_t *bfs_buf, __global oraclet_tmp_t *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 il_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
 {
   /**
    * base