X-Git-Url: https://git.sesse.net/?a=blobdiff_plain;f=src%2Ftt.cpp;h=e140a0ba10ec5fa2183b506b68750a4dcbc39289;hb=17c511922277e49eacdede3c92491ee32c303c30;hp=8ef2a63500acf02827028a6e485a2286881dfd6f;hpb=4251eac8601af47d1ee2d6f613f00a4bf00fedbb;p=stockfish

diff --git a/src/tt.cpp b/src/tt.cpp
index 8ef2a635..e140a0ba 100644
--- a/src/tt.cpp
+++ b/src/tt.cpp
@@ -30,10 +30,8 @@
 #include "movegen.h"
 #include "tt.h"
 
-
-/// This is the number of TTEntry slots for each position
-static const int ClusterSize = 5;
-
+// The main transposition table
+TranspositionTable TT;
 
 ////
 //// Functions
@@ -63,14 +61,14 @@ void TranspositionTable::set_size(unsigned mbSize) {
 
   // We store a cluster of ClusterSize number of TTEntry for each position
   // and newSize is the maximum number of storable positions.
-  while ((2 * newSize) * ClusterSize * (sizeof(TTEntry)) <= (mbSize << 20))
+  while ((2 * newSize) * sizeof(TTCluster) <= (mbSize << 20))
       newSize *= 2;
 
   if (newSize != size)
   {
       size = newSize;
       delete [] entries;
-      entries = new TTEntry[size * ClusterSize];
+      entries = new TTCluster[size];
       if (!entries)
       {
           std::cerr << "Failed to allocate " << mbSize
@@ -89,7 +87,7 @@ void TranspositionTable::set_size(unsigned mbSize) {
 
 void TranspositionTable::clear() {
 
-  memset(entries, 0, size * ClusterSize * sizeof(TTEntry));
+  memset(entries, 0, size * sizeof(TTCluster));
 }
 
 
@@ -99,7 +97,7 @@ void TranspositionTable::clear() {
 
 inline TTEntry* TranspositionTable::first_entry(const Key posKey) const {
 
-  return entries + ((uint32_t(posKey) & (size - 1)) * ClusterSize);
+  return entries[uint32_t(posKey) & (size - 1)].data;
 }
 
 
@@ -170,11 +168,19 @@ TTEntry* TranspositionTable::retrieve(const Key posKey) const {
 /// blocking function and do not stalls the CPU waiting for data
 /// to be loaded from RAM, that can be very slow. When we will
 /// subsequently call retrieve() the TT data will be already
-/// quickly accessible in L1/l2 CPU cache.
+/// quickly accessible in L1/L2 CPU cache.
 
 void TranspositionTable::prefetch(const Key posKey) const {
 
-  _mm_prefetch((char*)first_entry(posKey), _MM_HINT_T0);
+#if defined(__INTEL_COMPILER) || defined(__ICL)
+   // This hack prevents prefetches to be optimized away by the
+   // Intel compiler. Both MSVC and gcc seems not affected.
+   __asm__ ("");
+#endif
+
+   char const* addr = (char*)first_entry(posKey);
+  _mm_prefetch(addr, _MM_HINT_T2);
+  _mm_prefetch(addr+64, _MM_HINT_T2); // 64 bytes ahead
 }