Reuse 5 slots instead of 4

[stockfish] / src / tt.cpp

diff --git a/src/tt.cpp b/src/tt.cpp
index 8ef2a63500acf02827028a6e485a2286881dfd6f..5ea6a808acf86b2ad256f8715bf5c59772e015b1 100644 (file)
--- a/src/tt.cpp
+++ b/src/tt.cpp
@@ -25,15 +25,16 @@
 #include <cassert>
 #include <cmath>
 #include <cstring>
-#include <xmmintrin.h>
 
 #include "movegen.h"
 #include "tt.h"
 
+#if defined(_MSC_VER)
+#include <xmmintrin.h>
+#endif
 
-/// This is the number of TTEntry slots for each position
-static const int ClusterSize = 5;
-
+// The main transposition table
+TranspositionTable TT;
 
 ////
 //// Functions
@@ -63,14 +64,14 @@ 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 +90,7 @@ void TranspositionTable::set_size(unsigned mbSize) {
 
 void TranspositionTable::clear() {
 
-  memset(entries, 0, size * ClusterSize * sizeof(TTEntry));
+  memset(entries, 0, size * sizeof(TTCluster));
 }
 
 
@@ -99,7 +100,7 @@ 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 +171,20 @@ 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(_MSC_VER)
+   char* addr = (char*)first_entry(posKey);
+  _mm_prefetch(addr, _MM_HINT_T0);
+  _mm_prefetch(addr+64, _MM_HINT_T0);
+#else
+  // We need to force an asm volatile here because gcc builtin
+  // is optimized away by Intel compiler.
+  char* addr = (char*)first_entry(posKey);
+  asm volatile("prefetcht0 %0" :: "m" (addr));
+#endif
 }