Prepare for a more dynamic #pragma unroll use

OpenCL/kernel_serpent256.c

@@ -403,13 +403,17 @@
 
 void serpent256_set_key (u32 *ks, const u32 *ukey)
 {
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (int i = 0; i < 8; i++)
   {
     ks[i] = ukey[i];
   }
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (int i = 0; i < 132; i++)
   {
     ks[i + 8] = rotl32 (ks[i + 7] ^ ks[i + 5] ^ ks[i + 3] ^ ks[i + 0] ^ 0x9e3779b9 ^ i, 11);

OpenCL/m01450_a0.cl

@@ -112,7 +112,9 @@ void sha256_transform (const u32x w0[4], const u32x w1[4], const u32x w2[4], con
 
   ROUND_STEP (0);
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (int i = 16; i < 64; i += 16)
   {
     ROUND_EXPAND (); ROUND_STEP (i);

OpenCL/m01450_a1.cl

@@ -110,7 +110,9 @@ void sha256_transform (const u32x w0[4], const u32x w1[4], const u32x w2[4], con
 
   ROUND_STEP (0);
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (int i = 16; i < 64; i += 16)
   {
     ROUND_EXPAND (); ROUND_STEP (i);

OpenCL/m01450_a3.cl

@@ -110,7 +110,9 @@ void sha256_transform (const u32x w0[4], const u32x w1[4], const u32x w2[4], con
 
   ROUND_STEP (0);
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (int i = 16; i < 64; i += 16)
   {
     ROUND_EXPAND (); ROUND_STEP (i);

OpenCL/m01460_a0.cl

@@ -112,7 +112,9 @@ void sha256_transform (const u32x w0[4], const u32x w1[4], const u32x w2[4], con
 
   ROUND_STEP (0);
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (int i = 16; i < 64; i += 16)
   {
     ROUND_EXPAND (); ROUND_STEP (i);

OpenCL/m01460_a1.cl

@@ -110,7 +110,9 @@ void sha256_transform (const u32x w0[4], const u32x w1[4], const u32x w2[4], con
 
   ROUND_STEP (0);
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (int i = 16; i < 64; i += 16)
   {
     ROUND_EXPAND (); ROUND_STEP (i);

OpenCL/m01460_a3.cl

@@ -110,7 +110,9 @@ void sha256_transform (const u32x w0[4], const u32x w1[4], const u32x w2[4], con
 
   ROUND_STEP (0);
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (int i = 16; i < 64; i += 16)
   {
     ROUND_EXPAND (); ROUND_STEP (i);

OpenCL/m01500_a0.cl

@@ -371,7 +371,9 @@ void _des_crypt_keysetup (u32x c, u32x d, u32x Kc[16], u32x Kd[16], __local u32
 
   c = c & 0x0fffffff;
 
-  #pragma unroll 16
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (u32 i = 0; i < 16; i++)
   {
     if ((i < 2) || (i == 8) || (i == 15))
@@ -430,6 +432,9 @@ void _des_crypt_encrypt (u32x iv[2], u32 mask, u32x Kc[16], u32x Kd[16], __local
 
   for (u32 i = 0; i < 25; i++)
   {
+    #ifdef _unroll
+    #pragma unroll
+    #endif
     for (u32 j = 0; j < 16; j += 2)
     {
       u32x t;

OpenCL/m01500_a1.cl

@@ -368,7 +368,9 @@ void _des_crypt_keysetup (u32x c, u32x d, u32x Kc[16], u32x Kd[16], __local u32
 
   c = c & 0x0fffffff;
 
-  #pragma unroll 16
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (u32 i = 0; i < 16; i++)
   {
     if ((i < 2) || (i == 8) || (i == 15))
@@ -427,6 +429,9 @@ void _des_crypt_encrypt (u32x iv[2], u32 mask, u32x Kc[16], u32x Kd[16], __local
 
   for (u32 i = 0; i < 25; i++)
   {
+    #ifdef _unroll
+    #pragma unroll
+    #endif
     for (u32 j = 0; j < 16; j += 2)
     {
       u32x t;

OpenCL/m01500_a3.cl

@@ -1569,17 +1569,9 @@ void DESCrypt (const u32 SALT, const u32 K00, const u32 K01, const u32 K02, cons
 
   for (u32 ii = 0; ii < 25; ii++)
   {
-    #ifdef IS_NV
+    #ifdef _unroll
-    #if CUDA_ARCH >= 500
-    #else
     #pragma unroll
     #endif
-    #endif
-
-    #ifdef IS_AMD
-    #pragma unroll
-    #endif
-
     for (u32 i = 0; i < 2; i++)
     {
       if (i) KEYSET10 else KEYSET00
@@ -1705,17 +1697,9 @@ void DESCrypt (const u32 SALT, const u32 K00, const u32 K01, const u32 K02, cons
 
   for (u32 ii = 0; ii < 25; ii++)
   {
-    #ifdef IS_NV
+    #ifdef _unroll
-    #if CUDA_ARCH >= 500
-    #else
     #pragma unroll
     #endif
-    #endif
-
-    #ifdef IS_AMD
-    #pragma unroll
-    #endif
-
     for (u32 i = 0; i < 2; i++)
     {
       if (i) KEYSET10 else KEYSET00
@@ -2222,7 +2206,9 @@ void m01500m (__global pw_t *pws, __global kernel_rule_t *rules_buf, __global co
 
       u32 tmpResult = 0;
 
+      #ifdef _unroll
       #pragma unroll
+      #endif
       for (int i = 0; i < 32; i++)
       {
         const u32 b0 = -((search[0] >> i) & 1);
@@ -2249,7 +2235,9 @@ void m01500m (__global pw_t *pws, __global kernel_rule_t *rules_buf, __global co
     u32 out0[32];
     u32 out1[32];
 
+    #ifdef _unroll
     #pragma unroll
+    #endif
     for (int i = 0; i < 32; i++)
     {
       out0[i] = out[ 0 + 31 - i];
@@ -2259,7 +2247,9 @@ void m01500m (__global pw_t *pws, __global kernel_rule_t *rules_buf, __global co
     transpose32c (out0);
     transpose32c (out1);
 
+    #ifdef _unroll
     #pragma unroll
+    #endif
     for (int slice = 0; slice < 32; slice++)
     {
       const u32 r0 = out0[31 - slice];
@@ -2676,7 +2666,9 @@ __kernel void m01500_tm (__global u32 *mod, __global bs_word_t *words_buf_r)
 
   const u32 w0s = (w0 << 1) & 0xfefefefe;
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (int i = 0, j = 0; i < 32; i += 8, j += 7)
   {
     atomic_or (&words_buf_r[block].b[j + 0], (((w0s >> (i + 7)) & 1) << slice));

OpenCL/m01700_a0.cl

@@ -116,7 +116,9 @@ void sha512_transform (const u32x w0[4], const u32x w1[4], const u32x w2[4], con
 
   ROUND_STEP (0);
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (int i = 16; i < 80; i += 16)
   {
     ROUND_EXPAND (); ROUND_STEP (i);

OpenCL/m01700_a1.cl

@@ -114,7 +114,9 @@ void sha512_transform (const u32x w0[4], const u32x w1[4], const u32x w2[4], con
 
   ROUND_STEP (0);
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (int i = 16; i < 80; i += 16)
   {
     ROUND_EXPAND (); ROUND_STEP (i);

OpenCL/m01700_a3.cl

@@ -114,7 +114,9 @@ void sha512_transform (const u32x w0[4], const u32x w1[4], const u32x w2[4], con
 
   ROUND_STEP (0);
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (int i = 16; i < 80; i += 16)
   {
     ROUND_EXPAND (); ROUND_STEP (i);

OpenCL/m01710_a0.cl

@@ -116,7 +116,9 @@ void sha512_transform (const u32x w0[4], const u32x w1[4], const u32x w2[4], con
 
   ROUND_STEP (0);
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (int i = 16; i < 80; i += 16)
   {
     ROUND_EXPAND (); ROUND_STEP (i);

OpenCL/m01710_a1.cl

@@ -114,7 +114,9 @@ void sha512_transform (const u32x w0[4], const u32x w1[4], const u32x w2[4], con
 
   ROUND_STEP (0);
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (int i = 16; i < 80; i += 16)
   {
     ROUND_EXPAND (); ROUND_STEP (i);

OpenCL/m01710_a3.cl

@@ -114,7 +114,9 @@ void sha512_transform (const u32x w0[4], const u32x w1[4], const u32x w2[4], con
 
   ROUND_STEP (0);
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (int i = 16; i < 80; i += 16)
   {
     ROUND_EXPAND (); ROUND_STEP (i);

OpenCL/m01720_a0.cl

@@ -116,7 +116,9 @@ void sha512_transform (const u32x w0[4], const u32x w1[4], const u32x w2[4], con
 
   ROUND_STEP (0);
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (int i = 16; i < 80; i += 16)
   {
     ROUND_EXPAND (); ROUND_STEP (i);

OpenCL/m01720_a1.cl

@@ -114,7 +114,9 @@ void sha512_transform (const u32x w0[4], const u32x w1[4], const u32x w2[4], con
 
   ROUND_STEP (0);
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (int i = 16; i < 80; i += 16)
   {
     ROUND_EXPAND (); ROUND_STEP (i);

OpenCL/m01720_a3.cl

@@ -114,7 +114,9 @@ void sha512_transform (const u32x w0[4], const u32x w1[4], const u32x w2[4], con
 
   ROUND_STEP (0);
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (int i = 16; i < 80; i += 16)
   {
     ROUND_EXPAND (); ROUND_STEP (i);

OpenCL/m01730_a0.cl

@@ -116,7 +116,9 @@ void sha512_transform (const u32x w0[4], const u32x w1[4], const u32x w2[4], con
 
   ROUND_STEP (0);
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (int i = 16; i < 80; i += 16)
   {
     ROUND_EXPAND (); ROUND_STEP (i);

OpenCL/m01730_a1.cl

@@ -114,7 +114,9 @@ void sha512_transform (const u32x w0[4], const u32x w1[4], const u32x w2[4], con
 
   ROUND_STEP (0);
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (int i = 16; i < 80; i += 16)
   {
     ROUND_EXPAND (); ROUND_STEP (i);

OpenCL/m01730_a3.cl

@@ -114,7 +114,9 @@ void sha512_transform (const u32x w0[4], const u32x w1[4], const u32x w2[4], con
 
   ROUND_STEP (0);
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (int i = 16; i < 80; i += 16)
   {
     ROUND_EXPAND (); ROUND_STEP (i);

OpenCL/m01740_a0.cl

@@ -116,7 +116,9 @@ void sha512_transform (const u32x w0[4], const u32x w1[4], const u32x w2[4], con
 
   ROUND_STEP (0);
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (int i = 16; i < 80; i += 16)
   {
     ROUND_EXPAND (); ROUND_STEP (i);

OpenCL/m01740_a1.cl

@@ -114,7 +114,9 @@ void sha512_transform (const u32x w0[4], const u32x w1[4], const u32x w2[4], con
 
   ROUND_STEP (0);
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (int i = 16; i < 80; i += 16)
   {
     ROUND_EXPAND (); ROUND_STEP (i);

OpenCL/m01740_a3.cl

@@ -114,7 +114,9 @@ void sha512_transform (const u32x w0[4], const u32x w1[4], const u32x w2[4], con
 
   ROUND_STEP (0);
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (int i = 16; i < 80; i += 16)
   {
     ROUND_EXPAND (); ROUND_STEP (i);

OpenCL/m01750_a0.cl

@@ -116,7 +116,9 @@ void sha512_transform (const u64x w0[4], const u64x w1[4], const u64x w2[4], con
 
   ROUND_STEP (0);
 
-  //#pragma unroll
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (int i = 16; i < 80; i += 16)
   {
     ROUND_EXPAND (); ROUND_STEP (i);

OpenCL/m01750_a1.cl

@@ -114,7 +114,9 @@ void sha512_transform (const u64x w0[4], const u64x w1[4], const u64x w2[4], con
 
   ROUND_STEP (0);
 
-  //#pragma unroll
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (int i = 16; i < 80; i += 16)
   {
     ROUND_EXPAND (); ROUND_STEP (i);

OpenCL/m01750_a3.cl

@@ -114,7 +114,9 @@ void sha512_transform (const u64x w0[4], const u64x w1[4], const u64x w2[4], con
 
   ROUND_STEP (0);
 
-  //#pragma unroll
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (int i = 16; i < 80; i += 16)
   {
     ROUND_EXPAND (); ROUND_STEP (i);

OpenCL/m01760_a0.cl

@@ -116,7 +116,9 @@ void sha512_transform (const u64x w0[4], const u64x w1[4], const u64x w2[4], con
 
   ROUND_STEP (0);
 
-  //#pragma unroll
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (int i = 16; i < 80; i += 16)
   {
     ROUND_EXPAND (); ROUND_STEP (i);

OpenCL/m01760_a1.cl

@@ -114,7 +114,9 @@ void sha512_transform (const u64x w0[4], const u64x w1[4], const u64x w2[4], con
 
   ROUND_STEP (0);
 
-  //#pragma unroll
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (int i = 16; i < 80; i += 16)
   {
     ROUND_EXPAND (); ROUND_STEP (i);

OpenCL/m01760_a3.cl

@@ -114,7 +114,9 @@ void sha512_transform (const u64x w0[4], const u64x w1[4], const u64x w2[4], con
 
   ROUND_STEP (0);
 
-  //#pragma unroll
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (int i = 16; i < 80; i += 16)
   {
     ROUND_EXPAND (); ROUND_STEP (i);

OpenCL/m01800.cl

@@ -134,7 +134,9 @@ void sha512_transform (const u64 w[16], u64 digest[8])
 
   ROUND_STEP (0);
 
-  //#pragma unroll
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (int i = 16; i < 80; i += 16)
   {
     ROUND_EXPAND (); ROUND_STEP (i);
@@ -500,7 +502,9 @@ __kernel void m01800_loop (__global pw_t *pws, __global kernel_rule_t *rules_buf
     {
       const u32 block_len = wpc_len[pc];
 
-      #pragma unroll 64
+      #ifdef _unroll
+      #pragma unroll
+      #endif
       for (u32 k = 0, p = block_len - 64; k < 64; k++, p++)
       {
         PUTCHAR64_BE (block, p, GETCHAR64_BE (l_alt_result, k));

OpenCL/m03000_a0.cl

@@ -361,7 +361,9 @@ void _des_crypt_encrypt (u32x iv[2], u32x data[2], u32x Kc[16], u32x Kd[16], __l
   u32x r = data[0];
   u32x l = data[1];
 
-  #pragma unroll 16
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (u32 i = 0; i < 16; i += 2)
   {
     u32x u;
@@ -414,7 +416,9 @@ void _des_crypt_keysetup (u32x c, u32x d, u32x Kc[16], u32x Kd[16], __local u32
 
   c = c & 0x0fffffff;
 
-  #pragma unroll 16
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (u32 i = 0; i < 16; i++)
   {
     if ((i < 2) || (i == 8) || (i == 15))

OpenCL/m03000_a1.cl

@@ -358,7 +358,9 @@ void _des_crypt_encrypt (u32x iv[2], u32x data[2], u32x Kc[16], u32x Kd[16], __l
   u32x r = data[0];
   u32x l = data[1];
 
-  #pragma unroll 16
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (u32 i = 0; i < 16; i += 2)
   {
     u32x u;
@@ -411,7 +413,9 @@ void _des_crypt_keysetup (u32x c, u32x d, u32x Kc[16], u32x Kd[16], __local u32
 
   c = c & 0x0fffffff;
 
-  #pragma unroll 16
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (u32 i = 0; i < 16; i++)
   {
     if ((i < 2) || (i == 8) || (i == 15))

OpenCL/m03000_a3.cl

@@ -1546,21 +1546,9 @@ void DES (const u32 K00, const u32 K01, const u32 K02, const u32 K03, const u32
   KXX_DECL u32 k36, k37, k38, k39, k40, k41;
   KXX_DECL u32 k42, k43, k44, k45, k46, k47;
 
-  #ifdef IS_NV
+  #ifdef _unroll
-  #if CUDA_ARCH >= 500
-  #else
   #pragma unroll
   #endif
-  #endif
-
-  #ifdef IS_AMD
-  #pragma unroll
-  #endif
-
-  #ifdef IS_GENERIC
-  #pragma unroll 1
-  #endif
-
   for (u32 i = 0; i < 2; i++)
   {
     if (i) KEYSET10 else KEYSET00
@@ -2060,7 +2048,9 @@ void m03000m (__global pw_t *pws, __global kernel_rule_t *rules_buf, __global co
 
       u32 tmpResult = 0;
 
+      #ifdef _unroll
       #pragma unroll
+      #endif
       for (int i = 0; i < 32; i++)
       {
         const u32 b0 = -((search[0] >> i) & 1);
@@ -2087,7 +2077,9 @@ void m03000m (__global pw_t *pws, __global kernel_rule_t *rules_buf, __global co
     u32 out0[32];
     u32 out1[32];
 
+    #ifdef _unroll
     #pragma unroll
+    #endif
     for (int i = 0; i < 32; i++)
     {
       out0[i] = out[ 0 + 31 - i];
@@ -2097,7 +2089,9 @@ void m03000m (__global pw_t *pws, __global kernel_rule_t *rules_buf, __global co
     transpose32c (out0);
     transpose32c (out1);
 
+    #ifdef _unroll
     #pragma unroll
+    #endif
     for (int slice = 0; slice < 32; slice++)
     {
       const u32 r0 = out0[31 - slice];

OpenCL/m03100_a0.cl

@@ -384,7 +384,9 @@ void _des_crypt_encrypt (u32x iv[2], u32x data[2], u32x Kc[16], u32x Kd[16], __l
   r = rotl32 (r, 3u);
   l = rotl32 (l, 3u);
 
-  #pragma unroll 16
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (u32 i = 0; i < 16; i += 2)
   {
     u32x u;
@@ -442,7 +444,9 @@ void _des_crypt_keysetup (u32x c, u32x d, u32x Kc[16], u32x Kd[16], __local u32
 
   c = c & 0x0fffffff;
 
-  #pragma unroll 16
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (u32 i = 0; i < 16; i++)
   {
     if ((i < 2) || (i == 8) || (i == 15))

OpenCL/m03100_a1.cl

@@ -381,7 +381,9 @@ void _des_crypt_encrypt (u32x iv[2], u32x data[2], u32x Kc[16], u32x Kd[16], __l
   r = rotl32 (r, 3u);
   l = rotl32 (l, 3u);
 
-  #pragma unroll 16
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (u32 i = 0; i < 16; i += 2)
   {
     u32x u;
@@ -439,7 +441,9 @@ void _des_crypt_keysetup (u32x c, u32x d, u32x Kc[16], u32x Kd[16], __local u32
 
   c = c & 0x0fffffff;
 
-  #pragma unroll 16
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (u32 i = 0; i < 16; i++)
   {
     if ((i < 2) || (i == 8) || (i == 15))

OpenCL/m03100_a3.cl

@@ -381,7 +381,9 @@ void _des_crypt_encrypt (u32x iv[2], u32x data[2], u32x Kc[16], u32x Kd[16], __l
   r = rotl32 (r, 3u);
   l = rotl32 (l, 3u);
 
-  #pragma unroll 16
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (u32 i = 0; i < 16; i += 2)
   {
     u32x u;
@@ -439,7 +441,9 @@ void _des_crypt_keysetup (u32x c, u32x d, u32x Kc[16], u32x Kd[16], __local u32
 
   c = c & 0x0fffffff;
 
-  #pragma unroll 16
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (u32 i = 0; i < 16; i++)
   {
     if ((i < 2) || (i == 8) || (i == 15))

OpenCL/m03200.cl

@@ -679,7 +679,9 @@ __kernel void __attribute__((reqd_work_group_size (8, 1, 1))) m03200_loop (__glo
     L0 = 0;
     R0 = 0;
 
+    #ifdef _unroll
     #pragma unroll
+    #endif
     for (u32 i = 0; i < 9; i++)
     {
       BF_ENCRYPT (L0, R0);
@@ -742,7 +744,9 @@ __kernel void __attribute__((reqd_work_group_size (8, 1, 1))) m03200_loop (__glo
     L0 = 0;
     R0 = 0;
 
+    #ifdef _unroll
     #pragma unroll
+    #endif
     for (u32 i = 0; i < 9; i++)
     {
       BF_ENCRYPT (L0, R0);

OpenCL/m05200.cl

@@ -110,7 +110,9 @@ void sha256_transform (const u32 w0[4], const u32 w1[4], const u32 w2[4], const
 
   ROUND_STEP (0);
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (int i = 16; i < 64; i += 16)
   {
     ROUND_EXPAND (); ROUND_STEP (i);

OpenCL/m05500_a0.cl

@@ -358,7 +358,9 @@ void _des_crypt_encrypt (u32x iv[2], u32x data[2], u32x Kc[16], u32x Kd[16], __l
   u32x r = data[0];
   u32x l = data[1];
 
-  #pragma unroll 16
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (u32 i = 0; i < 16; i += 2)
   {
     u32x u;
@@ -411,7 +413,9 @@ void _des_crypt_keysetup (u32x c, u32x d, u32x Kc[16], u32x Kd[16], __local u32
 
   c = c & 0x0fffffff;
 
-  #pragma unroll 16
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (u32 i = 0; i < 16; i++)
   {
     if ((i < 2) || (i == 8) || (i == 15))

OpenCL/m05500_a1.cl

@@ -355,7 +355,9 @@ void _des_crypt_encrypt (u32x iv[2], u32x data[2], u32x Kc[16], u32x Kd[16], __l
   u32x r = data[0];
   u32x l = data[1];
 
-  #pragma unroll 16
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (u32 i = 0; i < 16; i += 2)
   {
     u32x u;
@@ -408,7 +410,9 @@ void _des_crypt_keysetup (u32x c, u32x d, u32x Kc[16], u32x Kd[16], __local u32
 
   c = c & 0x0fffffff;
 
-  #pragma unroll 16
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (u32 i = 0; i < 16; i++)
   {
     if ((i < 2) || (i == 8) || (i == 15))

OpenCL/m05500_a3.cl

@@ -356,7 +356,9 @@ void _des_crypt_encrypt (u32x iv[2], u32x data[2], u32x Kc[16], u32x Kd[16], __l
   u32x r = data[0];
   u32x l = data[1];
 
-  #pragma unroll 16
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (u32 i = 0; i < 16; i += 2)
   {
     u32x u;
@@ -409,7 +411,9 @@ void _des_crypt_keysetup (u32x c, u32x d, u32x Kc[16], u32x Kd[16], __local u32
 
   c = c & 0x0fffffff;
 
-  #pragma unroll 16
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (u32 i = 0; i < 16; i++)
   {
     if ((i < 2) || (i == 8) || (i == 15))

OpenCL/m06100_a0.cl

@@ -1180,7 +1180,9 @@ void whirlpool_transform (const u32x w[16], u32x dgst[16], __local u32 (*s_Ch)[2
   u32x Lh[8];
   u32x Ll[8];
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (int i = 0; i < 8; i++)
   {
     const u32x Lp0 = stateh[(i + 8) & 7] >> 24;
@@ -1233,7 +1235,9 @@ void whirlpool_transform (const u32x w[16], u32x dgst[16], __local u32 (*s_Ch)[2
     u32x Lh[8];
     u32x Ll[8];
 
+    #ifdef _unroll
     #pragma unroll
+    #endif
     for (int i = 0; i < 8; i++)
     {
       const u32x Lp0 = Kh[(i + 8) & 7] >> 24;
@@ -1281,7 +1285,9 @@ void whirlpool_transform (const u32x w[16], u32x dgst[16], __local u32 (*s_Ch)[2
     Kh[7] = Lh[7];
     Kl[7] = Ll[7];
 
-    #pragma unroll 8
+    #ifdef _unroll
+    #pragma unroll
+    #endif
     for (int i = 0; i < 8; i++)
     {
       const u32x Lp0 = stateh[(i + 8) & 7] >> 24;

OpenCL/m06100_a1.cl

@@ -1177,7 +1177,9 @@ void whirlpool_transform (const u32x w[16], u32x dgst[16], __local u32 (*s_Ch)[2
   u32x Lh[8];
   u32x Ll[8];
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (int i = 0; i < 8; i++)
   {
     const u32x Lp0 = stateh[(i + 8) & 7] >> 24;
@@ -1230,7 +1232,9 @@ void whirlpool_transform (const u32x w[16], u32x dgst[16], __local u32 (*s_Ch)[2
     u32x Lh[8];
     u32x Ll[8];
 
+    #ifdef _unroll
     #pragma unroll
+    #endif
     for (int i = 0; i < 8; i++)
     {
       const u32x Lp0 = Kh[(i + 8) & 7] >> 24;
@@ -1278,7 +1282,9 @@ void whirlpool_transform (const u32x w[16], u32x dgst[16], __local u32 (*s_Ch)[2
     Kh[7] = Lh[7];
     Kl[7] = Ll[7];
 
-    #pragma unroll 8
+    #ifdef _unroll
+    #pragma unroll
+    #endif
     for (int i = 0; i < 8; i++)
     {
       const u32x Lp0 = stateh[(i + 8) & 7] >> 24;

OpenCL/m06100_a3.cl

@@ -1178,7 +1178,9 @@ void whirlpool_transform (const u32x w[16], u32x dgst[16], __local u32 (*s_Ch)[2
   u32x Lh[8];
   u32x Ll[8];
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (int i = 0; i < 8; i++)
   {
     const u32x Lp0 = stateh[(i + 8) & 7] >> 24;
@@ -1231,7 +1233,9 @@ void whirlpool_transform (const u32x w[16], u32x dgst[16], __local u32 (*s_Ch)[2
     u32x Lh[8];
     u32x Ll[8];
 
+    #ifdef _unroll
     #pragma unroll
+    #endif
     for (int i = 0; i < 8; i++)
     {
       const u32x Lp0 = Kh[(i + 8) & 7] >> 24;
@@ -1279,7 +1283,9 @@ void whirlpool_transform (const u32x w[16], u32x dgst[16], __local u32 (*s_Ch)[2
     Kh[7] = Lh[7];
     Kl[7] = Ll[7];
 
-    #pragma unroll 8
+    #ifdef _unroll
+    #pragma unroll
+    #endif
     for (int i = 0; i < 8; i++)
     {
       const u32x Lp0 = stateh[(i + 8) & 7] >> 24;

OpenCL/m06221.cl

@@ -115,7 +115,9 @@ void sha512_transform (const u64 w[16], u64 dgst[8])
 
   ROUND_STEP (0);
 
-  //#pragma unroll
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (int i = 16; i < 80; i += 16)
   {
     ROUND_EXPAND (); ROUND_STEP (i);

OpenCL/m06222.cl

@@ -115,7 +115,9 @@ void sha512_transform (const u64 w[16], u64 dgst[8])
 
   ROUND_STEP (0);
 
-  //#pragma unroll
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (int i = 16; i < 80; i += 16)
   {
     ROUND_EXPAND (); ROUND_STEP (i);

OpenCL/m06223.cl

@@ -115,7 +115,9 @@ void sha512_transform (const u64 w[16], u64 dgst[8])
 
   ROUND_STEP (0);
 
-  //#pragma unroll
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (int i = 16; i < 80; i += 16)
   {
     ROUND_EXPAND (); ROUND_STEP (i);

OpenCL/m06231.cl

@@ -1175,7 +1175,9 @@ void whirlpool_transform_last (u32 dgst[16], __local u32 (*s_Ch)[256], __local u
 
     u32 i;
 
+    #ifdef _unroll
     #pragma unroll
+    #endif
     for (i = 0; i < 8; i++)
     {
       const u32 Lp0 = Kh[(i + 8) & 7] >> 24;
@@ -1223,7 +1225,9 @@ void whirlpool_transform_last (u32 dgst[16], __local u32 (*s_Ch)[256], __local u
     Kh[7] = Lh[7];
     Kl[7] = Ll[7];
 
+    #ifdef _unroll
     #pragma unroll
+    #endif
     for (i = 0; i < 8; i++)
     {
       const u32 Lp0 = stateh[(i + 8) & 7] >> 24;
@@ -1371,7 +1375,9 @@ void whirlpool_transform (const u32 w[16], u32 dgst[16], __local u32 (*s_Ch)[256
 
     u32 i;
 
+    #ifdef _unroll
     #pragma unroll
+    #endif
     for (i = 0; i < 8; i++)
     {
       const u32 Lp0 = Kh[(i + 8) & 7] >> 24;
@@ -1419,7 +1425,9 @@ void whirlpool_transform (const u32 w[16], u32 dgst[16], __local u32 (*s_Ch)[256
     Kh[7] = Lh[7];
     Kl[7] = Ll[7];
 
+    #ifdef _unroll
     #pragma unroll
+    #endif
     for (i = 0; i < 8; i++)
     {
       const u32 Lp0 = stateh[(i + 8) & 7] >> 24;

OpenCL/m06232.cl

@@ -1172,7 +1172,9 @@ void whirlpool_transform (const u32 w[16], u32 dgst[16], __local u32 (*s_Ch)[256
 
     u32 i;
 
-    #pragma unroll 8
+    #ifdef _unroll
+    #pragma unroll
+    #endif
     for (i = 0; i < 8; i++)
     {
       const u8 Lp0 = Kh[(i + 8) & 7] >> 24;
@@ -1220,7 +1222,9 @@ void whirlpool_transform (const u32 w[16], u32 dgst[16], __local u32 (*s_Ch)[256
     Kh[7] = Lh[7];
     Kl[7] = Ll[7];
 
-    #pragma unroll 8
+    #ifdef _unroll
+    #pragma unroll
+    #endif
     for (i = 0; i < 8; i++)
     {
       const u8 Lp0 = stateh[(i + 8) & 7] >> 24;

OpenCL/m06233.cl

@@ -1172,7 +1172,9 @@ void whirlpool_transform (const u32 w[16], u32 dgst[16], __local u32 (*s_Ch)[256
 
     u32 i;
 
-    #pragma unroll 8
+    #ifdef _unroll
+    #pragma unroll
+    #endif
     for (i = 0; i < 8; i++)
     {
       const u8 Lp0 = Kh[(i + 8) & 7] >> 24;
@@ -1220,7 +1222,9 @@ void whirlpool_transform (const u32 w[16], u32 dgst[16], __local u32 (*s_Ch)[256
     Kh[7] = Lh[7];
     Kl[7] = Ll[7];
 
-    #pragma unroll 8
+    #ifdef _unroll
+    #pragma unroll
+    #endif
     for (i = 0; i < 8; i++)
     {
       const u8 Lp0 = stateh[(i + 8) & 7] >> 24;

OpenCL/m06400.cl

@@ -110,7 +110,9 @@ void sha256_transform (const u32 w0[4], const u32 w1[4], const u32 w2[4], const
 
   ROUND_STEP (0);
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (int i = 16; i < 64; i += 16)
   {
     ROUND_EXPAND (); ROUND_STEP (i);

OpenCL/m06500.cl

@@ -114,7 +114,9 @@ void sha512_transform (const u64 w0[4], const u64 w1[4], const u64 w2[4], const
 
   ROUND_STEP (0);
 
-  //#pragma unroll
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (int i = 16; i < 80; i += 16)
   {
     ROUND_EXPAND (); ROUND_STEP (i);

OpenCL/m06600.cl

@@ -716,7 +716,9 @@ void AES128_ExpandKey (u32 *userkey, u32 *rek, __local u32 *s_te0, __local u32 *
   rek[2] = userkey[2];
   rek[3] = userkey[3];
 
-  #pragma unroll 10
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (u32 i = 0, j = 0; i < 10; i += 1, j += 4)
   {
     u32 temp = rek[j + 3];
@@ -1391,7 +1393,9 @@ __kernel void m06600_comp (__global pw_t *pws, __global kernel_rule_t *rules_buf
 
   AES128_ExpandKey (ukey, rek, s_te0, s_te1, s_te2, s_te3, s_te4);
 
-  #pragma unroll KEYLEN
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (u32 i = 0; i < KEYLEN; i++) rdk[i] = rek[i];
 
   AES128_InvertKey (rdk, s_td0, s_td1, s_td2, s_td3, s_td4, s_te0, s_te1, s_te2, s_te3, s_te4);

OpenCL/m06800.cl

@@ -1075,7 +1075,9 @@ void sha256_transform (const u32 w0[4], const u32 w1[4], const u32 w2[4], const
 
   ROUND_STEP (0);
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (int i = 16; i < 64; i += 16)
   {
     ROUND_EXPAND (); ROUND_STEP (i);
@@ -1541,7 +1543,9 @@ __kernel void m06800_comp (__global pw_t *pws, __global kernel_rule_t *rules_buf
 
   u32 rdk[KEYLEN];
 
-  #pragma unroll 60
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (u32 i = 0; i < KEYLEN; i++) rdk[i] = rek[i];
 
   AES256_InvertKey (rdk, s_td0, s_td1, s_td2, s_td3, s_td4, s_te0, s_te1, s_te2, s_te3, s_te4);

OpenCL/m07100.cl

@@ -114,7 +114,9 @@ void sha512_transform (const u64 w[16], u64 dgst[8])
 
   ROUND_STEP (0);
 
-  //#pragma unroll
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (int i = 16; i < 80; i += 16)
   {
     ROUND_EXPAND (); ROUND_STEP (i);

OpenCL/m07400.cl

@@ -112,7 +112,9 @@ void sha256_transform (const u32 w[16], u32 digest[8])
 
   ROUND_STEP (0);
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (int i = 16; i < 64; i += 16)
   {
     ROUND_EXPAND (); ROUND_STEP (i);
@@ -1303,7 +1305,9 @@ void sha256_transform (const u32 w[16], u32 digest[8])
 
   ROUND_STEP (0);
 
-  //#pragma unroll
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (int i = 16; i < 64; i += 16)
   {
     ROUND_EXPAND (); ROUND_STEP (i);
@@ -1686,7 +1690,9 @@ __kernel void m07400_loop (__global pw_t *pws, __global kernel_rule_t *rules_buf
 
     if (j1)
     {
-      #pragma unroll 32
+      #ifdef _unroll
+      #pragma unroll
+      #endif
       for (u32 k = 0, p = block_len - 32; k < 32; k++, p++)
       {
         PUTCHAR32_BE (block, p, GETCHAR32_BE (alt_result, k));

OpenCL/m07500_a0.cl

@@ -47,7 +47,9 @@ void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
 
   __local u32 *ptr = (__local u32 *) rc4_key->S;
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (u32 i = 0; i < 64; i++)
   {
     *ptr++ = v; v += a;
@@ -93,7 +95,9 @@ void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
 
 u8 rc4_next_16 (__local RC4_KEY *rc4_key, u8 i, u8 j, const u32 in[4], u32 out[4])
 {
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (u32 k = 0; k < 4; k++)
   {
     u32 xor4 = 0;

OpenCL/m07500_a1.cl

@@ -45,7 +45,9 @@ void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
 
   __local u32 *ptr = (__local u32 *) rc4_key->S;
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (u32 i = 0; i < 64; i++)
   {
     *ptr++ = v; v += a;
@@ -91,7 +93,9 @@ void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
 
 u8 rc4_next_16 (__local RC4_KEY *rc4_key, u8 i, u8 j, const u32 in[4], u32 out[4])
 {
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (u32 k = 0; k < 4; k++)
   {
     u32 xor4 = 0;

OpenCL/m07500_a3.cl

@@ -47,7 +47,9 @@ void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
 
   __local u32 *ptr = (__local u32 *) rc4_key->S;
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (u32 i = 0; i < 64; i++)
   {
     *ptr++ = v; v += a;
@@ -93,7 +95,9 @@ void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
 
 u8 rc4_next_16 (__local RC4_KEY *rc4_key, u8 i, u8 j, const u32 in[4], u32 out[4])
 {
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (u32 k = 0; k < 4; k++)
   {
     u32 xor4 = 0;

OpenCL/m07800_a0.cl

@@ -350,7 +350,9 @@ __kernel void m07800_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf,
     digest[3] = SHA1M_D;
     digest[4] = SHA1M_E;
 
-    #pragma unroll 32
+    #ifdef _unroll
+    #pragma unroll
+    #endif
     for (int i = 0; i < 32; i++) final[i] = 0;
 
     final[0] = w0[0];
@@ -594,7 +596,9 @@ __kernel void m07800_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf,
     digest[3] = SHA1M_D;
     digest[4] = SHA1M_E;
 
-    #pragma unroll 32
+    #ifdef _unroll
+    #pragma unroll
+    #endif
     for (int i = 0; i < 32; i++) final[i] = 0;
 
     final[0] = w0[0];

OpenCL/m07800_a1.cl

@@ -408,7 +408,9 @@ __kernel void m07800_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf,
     digest[3] = SHA1M_D;
     digest[4] = SHA1M_E;
 
-    #pragma unroll 32
+    #ifdef _unroll
+    #pragma unroll
+    #endif
     for (int i = 0; i < 32; i++) final[i] = 0;
 
     final[0] = w0[0];
@@ -712,7 +714,9 @@ __kernel void m07800_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf,
     digest[3] = SHA1M_D;
     digest[4] = SHA1M_E;
 
-    #pragma unroll 32
+    #ifdef _unroll
+    #pragma unroll
+    #endif
     for (int i = 0; i < 32; i++) final[i] = 0;
 
     final[0] = w0[0];

OpenCL/m07800_a3.cl

@@ -324,7 +324,9 @@ void m07800m (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], const u32 pw_len, __gl
     digest[3] = SHA1M_D;
     digest[4] = SHA1M_E;
 
-    #pragma unroll 32
+    #ifdef _unroll
+    #pragma unroll
+    #endif
     for (int i = 0; i < 32; i++) final[i] = 0;
 
     final[0] = swap32 (w0[0]);
@@ -536,7 +538,9 @@ void m07800s (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], const u32 pw_len, __gl
     digest[3] = SHA1M_D;
     digest[4] = SHA1M_E;
 
-    #pragma unroll 32
+    #ifdef _unroll
+    #pragma unroll
+    #endif
     for (int i = 0; i < 32; i++) final[i] = 0;
 
     final[0] = swap32 (w0[0]);

OpenCL/m07900.cl

@@ -114,7 +114,9 @@ void sha512_transform (const u64 w[16], u64 dgst[8])
 
   ROUND_STEP (0);
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (int i = 16; i < 80; i += 16)
   {
     ROUND_EXPAND (); ROUND_STEP (i);

OpenCL/m08000_a0.cl

@@ -119,7 +119,9 @@ void sha256_transform (u32x digest[8], const u32x w[16])
 
   ROUND_STEP (0);
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (int i = 16; i < 64; i += 16)
   {
     ROUND_EXPAND (); ROUND_STEP (i);
@@ -168,7 +170,9 @@ void sha256_transform_z (u32x digest[8])
 
   ROUND_STEP_Z (0);
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (int i = 16; i < 64; i += 16)
   {
     ROUND_STEP_Z (i);
@@ -217,7 +221,9 @@ void sha256_transform_s (u32x digest[8], __local u32 *w)
 
   ROUND_STEP_S (0);
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (int i = 16; i < 64; i += 16)
   {
     ROUND_STEP_S (i);
@@ -270,7 +276,9 @@ __kernel void m08000_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf,
   {
     w_s1[15] =               0 | salt_buf0 >> 16;
 
+    #ifdef _unroll
     #pragma unroll
+    #endif
     for (int i = 16; i < 64; i++)
     {
       w_s1[i] = SHA256_EXPAND_S (w_s1[i - 2], w_s1[i - 7], w_s1[i - 15], w_s1[i - 16]);
@@ -281,7 +289,9 @@ __kernel void m08000_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf,
     w_s2[ 2] = salt_buf2 << 16 | 0;
     w_s2[15] = (510 + 8) * 8;
 
+    #ifdef _unroll
     #pragma unroll
+    #endif
     for (int i = 16; i < 64; i++)
     {
       w_s2[i] = SHA256_EXPAND_S (w_s2[i - 2], w_s2[i - 7], w_s2[i - 15], w_s2[i - 16]);
@@ -437,7 +447,9 @@ __kernel void m08000_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf,
   {
     w_s1[15] =               0 | salt_buf0 >> 16;
 
+    #ifdef _unroll
     #pragma unroll
+    #endif
     for (int i = 16; i < 64; i++)
     {
       w_s1[i] = SHA256_EXPAND_S (w_s1[i - 2], w_s1[i - 7], w_s1[i - 15], w_s1[i - 16]);
@@ -448,7 +460,9 @@ __kernel void m08000_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf,
     w_s2[ 2] = salt_buf2 << 16 | 0;
     w_s2[15] = (510 + 8) * 8;
 
+    #ifdef _unroll
     #pragma unroll
+    #endif
     for (int i = 16; i < 64; i++)
     {
       w_s2[i] = SHA256_EXPAND_S (w_s2[i - 2], w_s2[i - 7], w_s2[i - 15], w_s2[i - 16]);

OpenCL/m08000_a1.cl

@@ -117,7 +117,9 @@ void sha256_transform (u32x digest[8], const u32x w[16])
 
   ROUND_STEP (0);
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (int i = 16; i < 64; i += 16)
   {
     ROUND_EXPAND (); ROUND_STEP (i);
@@ -166,7 +168,9 @@ void sha256_transform_z (u32x digest[8])
 
   ROUND_STEP_Z (0);
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (int i = 16; i < 64; i += 16)
   {
     ROUND_STEP_Z (i);
@@ -215,7 +219,9 @@ void sha256_transform_s (u32x digest[8], __local u32 *w)
 
   ROUND_STEP_S (0);
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (int i = 16; i < 64; i += 16)
   {
     ROUND_STEP_S (i);
@@ -268,7 +274,9 @@ __kernel void m08000_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf,
   {
     w_s1[15] =               0 | salt_buf0 >> 16;
 
+    #ifdef _unroll
     #pragma unroll
+    #endif
     for (int i = 16; i < 64; i++)
     {
       w_s1[i] = SHA256_EXPAND_S (w_s1[i - 2], w_s1[i - 7], w_s1[i - 15], w_s1[i - 16]);
@@ -279,7 +287,9 @@ __kernel void m08000_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf,
     w_s2[ 2] = salt_buf2 << 16 | 0;
     w_s2[15] = (510 + 8) * 8;
 
+    #ifdef _unroll
     #pragma unroll
+    #endif
     for (int i = 16; i < 64; i++)
     {
       w_s2[i] = SHA256_EXPAND_S (w_s2[i - 2], w_s2[i - 7], w_s2[i - 15], w_s2[i - 16]);
@@ -489,7 +499,9 @@ __kernel void m08000_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf,
   {
     w_s1[15] =               0 | salt_buf0 >> 16;
 
+    #ifdef _unroll
     #pragma unroll
+    #endif
     for (int i = 16; i < 64; i++)
     {
       w_s1[i] = SHA256_EXPAND_S (w_s1[i - 2], w_s1[i - 7], w_s1[i - 15], w_s1[i - 16]);
@@ -500,7 +512,9 @@ __kernel void m08000_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf,
     w_s2[ 2] = salt_buf2 << 16 | 0;
     w_s2[15] = (510 + 8) * 8;
 
+    #ifdef _unroll
     #pragma unroll
+    #endif
     for (int i = 16; i < 64; i++)
     {
       w_s2[i] = SHA256_EXPAND_S (w_s2[i - 2], w_s2[i - 7], w_s2[i - 15], w_s2[i - 16]);

OpenCL/m08000_a3.cl

@@ -117,7 +117,9 @@ void sha256_transform (u32x digest[8], const u32x w[16])
 
   ROUND_STEP (0);
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (int i = 16; i < 64; i += 16)
   {
     ROUND_EXPAND (); ROUND_STEP (i);
@@ -166,7 +168,9 @@ void sha256_transform_z (u32x digest[8])
 
   ROUND_STEP_Z (0);
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (int i = 16; i < 64; i += 16)
   {
     ROUND_STEP_Z (i);
@@ -215,7 +219,9 @@ void sha256_transform_s (u32x digest[8], __local u32 *w)
 
   ROUND_STEP_S (0);
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (int i = 16; i < 64; i += 16)
   {
     ROUND_STEP_S (i);
@@ -265,7 +271,9 @@ void m08000m (__local u32 *w_s1, __local u32 *w_s2, u32 w[16], const u32 pw_len,
   {
     w_s1[15] =               0 | salt_buf0 >> 16;
 
+    #ifdef _unroll
     #pragma unroll
+    #endif
     for (int i = 16; i < 64; i++)
     {
       w_s1[i] = SHA256_EXPAND_S (w_s1[i - 2], w_s1[i - 7], w_s1[i - 15], w_s1[i - 16]);
@@ -276,7 +284,9 @@ void m08000m (__local u32 *w_s1, __local u32 *w_s2, u32 w[16], const u32 pw_len,
     w_s2[ 2] = salt_buf2 << 16 | 0;
     w_s2[15] = (510 + 8) * 8;
 
+    #ifdef _unroll
     #pragma unroll
+    #endif
     for (int i = 16; i < 64; i++)
     {
       w_s2[i] = SHA256_EXPAND_S (w_s2[i - 2], w_s2[i - 7], w_s2[i - 15], w_s2[i - 16]);
@@ -383,7 +393,9 @@ void m08000s (__local u32 *w_s1, __local u32 *w_s2, u32 w[16], const u32 pw_len,
   {
     w_s1[15] =               0 | salt_buf0 >> 16;
 
+    #ifdef _unroll
     #pragma unroll
+    #endif
     for (int i = 16; i < 64; i++)
     {
       w_s1[i] = SHA256_EXPAND_S (w_s1[i - 2], w_s1[i - 7], w_s1[i - 15], w_s1[i - 16]);
@@ -394,7 +406,9 @@ void m08000s (__local u32 *w_s1, __local u32 *w_s2, u32 w[16], const u32 pw_len,
     w_s2[ 2] = salt_buf2 << 16 | 0;
     w_s2[15] = (510 + 8) * 8;
 
+    #ifdef _unroll
     #pragma unroll
+    #endif
     for (int i = 16; i < 64; i++)
     {
       w_s2[i] = SHA256_EXPAND_S (w_s2[i - 2], w_s2[i - 7], w_s2[i - 15], w_s2[i - 16]);

OpenCL/m08200.cl

@@ -134,7 +134,9 @@ void sha256_transform (const u32 w0[4], const u32 w1[4], const u32 w2[4], const
 
   ROUND_STEP (0);
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (int i = 16; i < 64; i += 16)
   {
     ROUND_EXPAND (); ROUND_STEP (i);
@@ -321,7 +323,9 @@ void sha512_transform (const u64 w[16], u64 dgst[8])
 
   ROUND512_STEP (0);
 
-  //#pragma unroll
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (int i = 16; i < 80; i += 16)
   {
     ROUND512_EXPAND (); ROUND512_STEP (i);

OpenCL/m08500_a0.cl

@@ -410,7 +410,9 @@ void _des_crypt_encrypt (u32x iv[2], u32x data[2], u32x Kc[16], u32x Kd[16], __l
   u32x r = data[0];
   u32x l = data[1];
 
-  #pragma unroll 16
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (u32 i = 0; i < 16; i += 2)
   {
     u32x u;
@@ -463,7 +465,9 @@ void _des_crypt_keysetup (u32x c, u32x d, u32x Kc[16], u32x Kd[16], __local u32
 
   c = c & 0x0fffffff;
 
-  #pragma unroll 16
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (u32 i = 0; i < 16; i++)
   {
     if ((i < 2) || (i == 8) || (i == 15))

OpenCL/m08500_a1.cl

@@ -407,7 +407,9 @@ void _des_crypt_encrypt (u32x iv[2], u32x data[2], u32x Kc[16], u32x Kd[16], __l
   u32x r = data[0];
   u32x l = data[1];
 
-  #pragma unroll 16
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (u32 i = 0; i < 16; i += 2)
   {
     u32x u;
@@ -460,7 +462,9 @@ void _des_crypt_keysetup (u32x c, u32x d, u32x Kc[16], u32x Kd[16], __local u32
 
   c = c & 0x0fffffff;
 
-  #pragma unroll 16
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (u32 i = 0; i < 16; i++)
   {
     if ((i < 2) || (i == 8) || (i == 15))

OpenCL/m08500_a3.cl

@@ -408,7 +408,9 @@ void _des_crypt_encrypt (u32x iv[2], u32x data[2], u32x Kc[16], u32x Kd[16], __l
   u32x r = data[0];
   u32x l = data[1];
 
-  #pragma unroll 16
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (u32 i = 0; i < 16; i += 2)
   {
     u32x u;
@@ -461,7 +463,9 @@ void _des_crypt_keysetup (u32x c, u32x d, u32x Kc[16], u32x Kd[16], __local u32
 
   c = c & 0x0fffffff;
 
-  #pragma unroll 16
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (u32 i = 0; i < 16; i++)
   {
     if ((i < 2) || (i == 8) || (i == 15))

OpenCL/m08600_a0.cl

@@ -82,7 +82,9 @@ void lotus_mix (u32x *in, __local u32 *s_lotus_magic_table)
   {
     u32 s = 48;
 
-    #pragma unroll 12
+    #ifdef _unroll
+    #pragma unroll
+    #endif
     for (int j = 0; j < 12; j++)
     {
       u32x tmp_in = in[j];
@@ -104,7 +106,9 @@ void lotus_transform_password (u32x in[4], u32x out[4], __local u32 *s_lotus_mag
 
   u32x c;
 
-  #pragma unroll 4
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (int i = 0; i < 4; i++)
   {
     t ^= (in[i] >>  0) & 0xff; c = BOX1 (s_lotus_magic_table, t); out[i] ^= c <<  0; t = ((out[i] >>  0) & 0xff);

OpenCL/m08600_a1.cl

@@ -79,7 +79,9 @@ void lotus_mix (u32x *in, __local u32 *s_lotus_magic_table)
   {
     u32 s = 48;
 
-    #pragma unroll 12
+    #ifdef _unroll
+    #pragma unroll
+    #endif
     for (int j = 0; j < 12; j++)
     {
       u32x tmp_in = in[j];
@@ -101,7 +103,9 @@ void lotus_transform_password (u32x in[4], u32x out[4], __local u32 *s_lotus_mag
 
   u32x c;
 
-  #pragma unroll 4
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (int i = 0; i < 4; i++)
   {
     t ^= (in[i] >>  0) & 0xff; c = BOX1 (s_lotus_magic_table, t); out[i] ^= c <<  0; t = ((out[i] >>  0) & 0xff);

OpenCL/m08600_a3.cl

@@ -79,7 +79,9 @@ void lotus_mix (u32x *in, __local u32 *s_lotus_magic_table)
   {
     u32 s = 48;
 
-    #pragma unroll 12
+    #ifdef _unroll
+    #pragma unroll
+    #endif
     for (int j = 0; j < 12; j++)
     {
       u32x tmp_in = in[j];
@@ -101,7 +103,9 @@ void lotus_transform_password (u32x in[4], u32x out[4], __local u32 *s_lotus_mag
 
   u32x c;
 
-  #pragma unroll 4
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (int i = 0; i < 4; i++)
   {
     t ^= (in[i] >>  0) & 0xff; c = BOX1 (s_lotus_magic_table, t); out[i] ^= c <<  0; t = ((out[i] >>  0) & 0xff);

OpenCL/m08700_a0.cl

@@ -94,7 +94,9 @@ void lotus_mix (u32x *in, __local u32 *s_lotus_magic_table)
   {
     u32 s = 48;
 
-    #pragma unroll 12
+    #ifdef _unroll
+    #pragma unroll
+    #endif
     for (int j = 0; j < 12; j++)
     {
       u32x tmp_in = in[j];
@@ -116,7 +118,9 @@ void lotus_transform_password (u32x in[4], u32x out[4], __local u32 *s_lotus_mag
 
   u32x c;
 
-  #pragma unroll 4
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (int i = 0; i < 4; i++)
   {
     t ^= (in[i] >>  0) & 0xff; c = BOX1 (s_lotus_magic_table, t); out[i] ^= c <<  0; t = ((out[i] >>  0) & 0xff);

OpenCL/m08700_a1.cl

@@ -91,7 +91,9 @@ void lotus_mix (u32x *in, __local u32 *s_lotus_magic_table)
   {
     u32 s = 48;
 
-    #pragma unroll 12
+    #ifdef _unroll
+    #pragma unroll
+    #endif
     for (int j = 0; j < 12; j++)
     {
       u32x tmp_in = in[j];
@@ -113,7 +115,9 @@ void lotus_transform_password (u32x in[4], u32x out[4], __local u32 *s_lotus_mag
 
   u32x c;
 
-  #pragma unroll 4
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (int i = 0; i < 4; i++)
   {
     t ^= (in[i] >>  0) & 0xff; c = BOX1 (s_lotus_magic_table, t); out[i] ^= c <<  0; t = ((out[i] >>  0) & 0xff);

OpenCL/m08700_a3.cl

@@ -93,7 +93,9 @@ void lotus_mix (u32x *in, __local u32 *s_lotus_magic_table)
   {
     u32 s = 48;
 
-    #pragma unroll 12
+    #ifdef _unroll
+    #pragma unroll
+    #endif
     for (int j = 0; j < 12; j++)
     {
       u32x tmp_in = in[j];
@@ -115,7 +117,9 @@ void lotus_transform_password (u32x in[4], u32x out[4], __local u32 *s_lotus_mag
 
   u32x c;
 
-  #pragma unroll 4
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (int i = 0; i < 4; i++)
   {
     t ^= (in[i] >>  0) & 0xff; c = BOX1 (s_lotus_magic_table, t); out[i] ^= c <<  0; t = ((out[i] >>  0) & 0xff);

OpenCL/m08800.cl

@@ -1212,7 +1212,9 @@ void sha256_transform (const u32 w0[4], const u32 w1[4], const u32 w2[4], const
 
   ROUND_STEP (0);
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (int i = 16; i < 64; i += 16)
   {
     ROUND_EXPAND (); ROUND_STEP (i);

OpenCL/m08900.cl

@@ -110,7 +110,9 @@ void sha256_transform (const u32 w0[4], const u32 w1[4], const u32 w2[4], const
 
   ROUND_STEP (0);
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (int i = 16; i < 64; i += 16)
   {
     ROUND_EXPAND (); ROUND_STEP (i);
@@ -705,7 +707,9 @@ void scrypt_smix (uint4 *X, uint4 *T, const u32 phy, __global uint4 *V)
 
   const u32 x = gid % xSIZE;
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (u32 i = 0; i < STATE_CNT4; i += 4)
   {
     T[0] = (uint4) (X[i + 0].x, X[i + 1].y, X[i + 2].z, X[i + 3].w);
@@ -743,7 +747,9 @@ void scrypt_smix (uint4 *X, uint4 *T, const u32 phy, __global uint4 *V)
     salsa_r (X);
   }
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (u32 i = 0; i < STATE_CNT4; i += 4)
   {
     T[0] = (uint4) (X[i + 0].x, X[i + 3].y, X[i + 2].z, X[i + 1].w);
@@ -910,12 +916,16 @@ __kernel void m08900_loop (__global pw_t *pws, __global kernel_rule_t *rules_buf
   uint4 X[STATE_CNT4];
   uint4 T[STATE_CNT4];
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (int z = 0; z < STATE_CNT4; z++) X[z] = swap32_4 (tmps[gid].P[z]);
 
   scrypt_smix (X, T, scrypt_phy, d_scryptV_buf);
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (int z = 0; z < STATE_CNT4; z++) tmps[gid].P[z] = swap32_4 (X[z]);
 
   #if SCRYPT_P >= 1

OpenCL/m09100.cl

@@ -75,7 +75,9 @@ void lotus_mix (u32 *in, __local u32 *s_lotus_magic_table)
   {
     u32 s = 48;
 
-    #pragma unroll 12
+    #ifdef _unroll
+    #pragma unroll
+    #endif
     for (int j = 0; j < 12; j++)
     {
       u32 tmp_in = in[j];
@@ -97,7 +99,9 @@ void lotus_transform_password (u32 in[4], u32 out[4], __local u32 *s_lotus_magic
 
   u32 c;
 
-  #pragma unroll 4
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (int i = 0; i < 4; i++)
   {
     t ^= (in[i] >>  0) & 0xff; c = BOX (s_lotus_magic_table, t); out[i] ^= c <<  0; t = ((out[i] >>  0) & 0xff);

OpenCL/m09600.cl

@@ -1073,10 +1073,7 @@ void sha512_transform (const u64 w0[4], const u64 w1[4], const u64 w2[4], const
 
   ROUND_STEP (0);
 
-  #ifdef IS_AMD
+  #ifdef _unroll
-  // #pragma unroll
-  // breaks compiler
-  #else
   #pragma unroll
   #endif
   for (int i = 16; i < 80; i += 16)

OpenCL/m09700_a0.cl

@@ -47,7 +47,9 @@ void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
 
   __local u32 *ptr = (__local u32 *) rc4_key->S;
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (u32 i = 0; i < 64; i++)
   {
     *ptr++ = v; v += a;
@@ -93,7 +95,9 @@ void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
 
 u8 rc4_next_16 (__local RC4_KEY *rc4_key, u8 i, u8 j, const u32 in[4], u32 out[4])
 {
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (u32 k = 0; k < 4; k++)
   {
     u32 xor4 = 0;

OpenCL/m09700_a1.cl

@@ -45,7 +45,9 @@ void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
 
   __local u32 *ptr = (__local u32 *) rc4_key->S;
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (u32 i = 0; i < 64; i++)
   {
     *ptr++ = v; v += a;
@@ -91,7 +93,9 @@ void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
 
 u8 rc4_next_16 (__local RC4_KEY *rc4_key, u8 i, u8 j, const u32 in[4], u32 out[4])
 {
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (u32 k = 0; k < 4; k++)
   {
     u32 xor4 = 0;

OpenCL/m09700_a3.cl

@@ -44,7 +44,9 @@ void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
 
   __local u32 *ptr = (__local u32 *) rc4_key->S;
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (u32 i = 0; i < 64; i++)
   {
     *ptr++ = v; v += a;
@@ -90,7 +92,9 @@ void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
 
 u8 rc4_next_16 (__local RC4_KEY *rc4_key, u8 i, u8 j, const u32 in[4], u32 out[4])
 {
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (u32 k = 0; k < 4; k++)
   {
     u32 xor4 = 0;

OpenCL/m09710_a0.cl

@@ -47,7 +47,9 @@ void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
 
   __local u32 *ptr = (__local u32 *) rc4_key->S;
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (u32 i = 0; i < 64; i++)
   {
     *ptr++ = v; v += a;
@@ -93,7 +95,9 @@ void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
 
 u8 rc4_next_16 (__local RC4_KEY *rc4_key, u8 i, u8 j, const u32 in[4], u32 out[4])
 {
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (u32 k = 0; k < 4; k++)
   {
     u32 xor4 = 0;

OpenCL/m09710_a1.cl

@@ -45,7 +45,9 @@ void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
 
   __local u32 *ptr = (__local u32 *) rc4_key->S;
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (u32 i = 0; i < 64; i++)
   {
     *ptr++ = v; v += a;
@@ -91,7 +93,9 @@ void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
 
 u8 rc4_next_16 (__local RC4_KEY *rc4_key, u8 i, u8 j, const u32 in[4], u32 out[4])
 {
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (u32 k = 0; k < 4; k++)
   {
     u32 xor4 = 0;

OpenCL/m09710_a3.cl

@@ -47,7 +47,9 @@ void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
 
   __local u32 *ptr = (__local u32 *) rc4_key->S;
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (u32 i = 0; i < 64; i++)
   {
     *ptr++ = v; v += a;
@@ -93,7 +95,9 @@ void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
 
 u8 rc4_next_16 (__local RC4_KEY *rc4_key, u8 i, u8 j, const u32 in[4], u32 out[4])
 {
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (u32 k = 0; k < 4; k++)
   {
     u32 xor4 = 0;

OpenCL/m09800_a0.cl

@@ -47,7 +47,9 @@ void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
 
   __local u32 *ptr = (__local u32 *) rc4_key->S;
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (u32 i = 0; i < 64; i++)
   {
     *ptr++ = v; v += a;
@@ -93,7 +95,9 @@ void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
 
 u8 rc4_next_16 (__local RC4_KEY *rc4_key, u8 i, u8 j, const u32 in[4], u32 out[4])
 {
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (u32 k = 0; k < 4; k++)
   {
     u32 xor4 = 0;

OpenCL/m09800_a1.cl

@@ -45,7 +45,9 @@ void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
 
   __local u32 *ptr = (__local u32 *) rc4_key->S;
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (u32 i = 0; i < 64; i++)
   {
     *ptr++ = v; v += a;
@@ -91,7 +93,9 @@ void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
 
 u8 rc4_next_16 (__local RC4_KEY *rc4_key, u8 i, u8 j, const u32 in[4], u32 out[4])
 {
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (u32 k = 0; k < 4; k++)
   {
     u32 xor4 = 0;

OpenCL/m09800_a3.cl

@@ -44,7 +44,9 @@ void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
 
   __local u32 *ptr = (__local u32 *) rc4_key->S;
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (u32 i = 0; i < 64; i++)
   {
     *ptr++ = v; v += a;
@@ -90,7 +92,9 @@ void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
 
 u8 rc4_next_16 (__local RC4_KEY *rc4_key, u8 i, u8 j, const u32 in[4], u32 out[4])
 {
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (u32 k = 0; k < 4; k++)
   {
     u32 xor4 = 0;

OpenCL/m09810_a0.cl

@@ -47,7 +47,9 @@ void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
 
   __local u32 *ptr = (__local u32 *) rc4_key->S;
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (u32 i = 0; i < 64; i++)
   {
     *ptr++ = v; v += a;
@@ -93,7 +95,9 @@ void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
 
 u8 rc4_next_16 (__local RC4_KEY *rc4_key, u8 i, u8 j, const u32 in[4], u32 out[4])
 {
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (u32 k = 0; k < 4; k++)
   {
     u32 xor4 = 0;

OpenCL/m09810_a1.cl

@@ -45,7 +45,9 @@ void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
 
   __local u32 *ptr = (__local u32 *) rc4_key->S;
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (u32 i = 0; i < 64; i++)
   {
     *ptr++ = v; v += a;
@@ -91,7 +93,9 @@ void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
 
 u8 rc4_next_16 (__local RC4_KEY *rc4_key, u8 i, u8 j, const u32 in[4], u32 out[4])
 {
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (u32 k = 0; k < 4; k++)
   {
     u32 xor4 = 0;

OpenCL/m09810_a3.cl

@@ -47,7 +47,9 @@ void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
 
   __local u32 *ptr = (__local u32 *) rc4_key->S;
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (u32 i = 0; i < 64; i++)
   {
     *ptr++ = v; v += a;
@@ -93,7 +95,9 @@ void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
 
 u8 rc4_next_16 (__local RC4_KEY *rc4_key, u8 i, u8 j, const u32 in[4], u32 out[4])
 {
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (u32 k = 0; k < 4; k++)
   {
     u32 xor4 = 0;

OpenCL/m10400_a0.cl

@@ -61,7 +61,9 @@ void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
 
   __local u32 *ptr = (__local u32 *) rc4_key->S;
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (u32 i = 0; i < 64; i++)
   {
     ptr[i] = v; v += a;
@@ -75,7 +77,9 @@ void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
 
   u32 j = 0;
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (u32 i = 0; i < 255; i += 5)
   {
     j += rc4_key->S[i + 0] + d0; swap (rc4_key, i + 0, j);
@@ -90,7 +94,9 @@ void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
 
 u8 rc4_next_16 (__local RC4_KEY *rc4_key, u8 i, u8 j, __constant u32 *in, u32 out[4])
 {
-  #pragma unroll 4
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (u32 k = 0; k < 4; k++)
   {
     u32 xor4 = 0;

OpenCL/m10400_a1.cl

@@ -59,7 +59,9 @@ void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
 
   __local u32 *ptr = (__local u32 *) rc4_key->S;
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (u32 i = 0; i < 64; i++)
   {
     ptr[i] = v; v += a;
@@ -73,7 +75,9 @@ void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
 
   u32 j = 0;
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (u32 i = 0; i < 255; i += 5)
   {
     j += rc4_key->S[i + 0] + d0; swap (rc4_key, i + 0, j);
@@ -88,7 +92,9 @@ void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
 
 u8 rc4_next_16 (__local RC4_KEY *rc4_key, u8 i, u8 j, __constant u32 *in, u32 out[4])
 {
-  #pragma unroll 4
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (u32 k = 0; k < 4; k++)
   {
     u32 xor4 = 0;

OpenCL/m10400_a3.cl

@@ -59,7 +59,9 @@ void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
 
   __local u32 *ptr = (__local u32 *) rc4_key->S;
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (u32 i = 0; i < 64; i++)
   {
     ptr[i] = v; v += a;
@@ -73,7 +75,9 @@ void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
 
   u32 j = 0;
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (u32 i = 0; i < 255; i += 5)
   {
     j += rc4_key->S[i + 0] + d0; swap (rc4_key, i + 0, j);
@@ -88,7 +92,9 @@ void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
 
 u8 rc4_next_16 (__local RC4_KEY *rc4_key, u8 i, u8 j, __constant u32 *in, u32 out[4])
 {
-  #pragma unroll 4
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (u32 k = 0; k < 4; k++)
   {
     u32 xor4 = 0;

OpenCL/m10410_a0.cl

@@ -61,7 +61,9 @@ void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
 
   __local u32 *ptr = (__local u32 *) rc4_key->S;
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (u32 i = 0; i < 64; i++)
   {
     ptr[i] = v; v += a;
@@ -75,7 +77,9 @@ void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
 
   u32 j = 0;
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (u32 i = 0; i < 255; i += 5)
   {
     j += rc4_key->S[i + 0] + d0; swap (rc4_key, i + 0, j);
@@ -90,7 +94,9 @@ void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
 
 u8 rc4_next_16 (__local RC4_KEY *rc4_key, u8 i, u8 j, __constant u32 *in, u32 out[4])
 {
-  #pragma unroll 4
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (u32 k = 0; k < 4; k++)
   {
     u32 xor4 = 0;

OpenCL/m10410_a1.cl

@@ -59,7 +59,9 @@ void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
 
   __local u32 *ptr = (__local u32 *) rc4_key->S;
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (u32 i = 0; i < 64; i++)
   {
     ptr[i] = v; v += a;
@@ -73,7 +75,9 @@ void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
 
   u32 j = 0;
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (u32 i = 0; i < 255; i += 5)
   {
     j += rc4_key->S[i + 0] + d0; swap (rc4_key, i + 0, j);
@@ -88,7 +92,9 @@ void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
 
 u8 rc4_next_16 (__local RC4_KEY *rc4_key, u8 i, u8 j, __constant u32 *in, u32 out[4])
 {
-  #pragma unroll 4
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (u32 k = 0; k < 4; k++)
   {
     u32 xor4 = 0;

OpenCL/m10410_a3.cl

@@ -59,7 +59,9 @@ void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
 
   __local u32 *ptr = (__local u32 *) rc4_key->S;
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (u32 i = 0; i < 64; i++)
   {
     ptr[i] = v; v += a;
@@ -73,7 +75,9 @@ void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
 
   u32 j = 0;
 
+  #ifdef _unroll
   #pragma unroll
+  #endif
   for (u32 i = 0; i < 255; i += 5)
   {
     j += rc4_key->S[i + 0] + d0; swap (rc4_key, i + 0, j);
@@ -88,7 +92,9 @@ void rc4_init_16 (__local RC4_KEY *rc4_key, const u32 data[4])
 
 u8 rc4_next_16 (__local RC4_KEY *rc4_key, u8 i, u8 j, __constant u32 *in, u32 out[4])
 {
-  #pragma unroll 4
+  #ifdef _unroll
+  #pragma unroll
+  #endif
   for (u32 k = 0; k < 4; k++)
   {
     u32 xor4 = 0;