Add support for Windows large pages

for users that set the needed privilige "Lock Pages in Memory"
large pages will be automatically enabled (see Readme.md).

This expert setting might improve speed, 5% - 30%, depending
on the hardware, the number of threads and hash size. More for
large hashes, large number of threads and NUMA. If the operating
system can not allocate large pages (easier after a reboot), default
allocation is used automatically. The engine log provides details.

closes https://github.com/official-stockfish/Stockfish/pull/2656

fixes https://github.com/official-stockfish/Stockfish/issues/2619

No functional change

src/main.cpp

@@ -49,6 +49,7 @@ int main(int argc, char* argv[]) {
 
   UCI::loop(argc, argv);
 
+  TT.resize(0);
   Threads.set(0);
   return 0;
 }

src/misc.cpp

@@ -309,6 +309,69 @@ void* aligned_ttmem_alloc(size_t allocSize, void*& mem) {
   return mem;
 }
 
+#elif defined(_WIN64)
+
+static void* aligned_ttmem_alloc_large_pages(size_t allocSize) {
+
+  HANDLE hProcessToken { };
+  LUID luid { };
+  void* mem = nullptr;
+
+  const size_t largePageSize = GetLargePageMinimum();
+  if (!largePageSize)
+      return nullptr;
+
+  // We need SeLockMemoryPrivilege, so try to enable it for the process
+  if (!OpenProcessToken(GetCurrentProcess(), TOKEN_ADJUST_PRIVILEGES | TOKEN_QUERY, &hProcessToken))
+      return nullptr;
+
+  if (LookupPrivilegeValue(NULL, SE_LOCK_MEMORY_NAME, &luid))
+  {
+      TOKEN_PRIVILEGES tp { };
+      TOKEN_PRIVILEGES prevTp { };
+      DWORD prevTpLen = 0;
+
+      tp.PrivilegeCount = 1;
+      tp.Privileges[0].Luid = luid;
+      tp.Privileges[0].Attributes = SE_PRIVILEGE_ENABLED;
+
+      // Try to enable SeLockMemoryPrivilege. Note that even if AdjustTokenPrivileges() succeeds,
+      // we still need to query GetLastError() to ensure that the privileges were actually obtained...
+      if (AdjustTokenPrivileges(
+              hProcessToken, FALSE, &tp, sizeof(TOKEN_PRIVILEGES), &prevTp, &prevTpLen) &&
+          GetLastError() == ERROR_SUCCESS)
+      {
+          // round up size to full pages and allocate
+          allocSize = (allocSize + largePageSize - 1) & ~size_t(largePageSize - 1);
+          mem = VirtualAlloc(
+              NULL, allocSize, MEM_RESERVE | MEM_COMMIT | MEM_LARGE_PAGES, PAGE_READWRITE);
+
+          // privilege no longer needed, restore previous state
+          AdjustTokenPrivileges(hProcessToken, FALSE, &prevTp, 0, NULL, NULL);
+      }
+  }
+
+  CloseHandle(hProcessToken);
+
+  return mem;
+}
+
+void* aligned_ttmem_alloc(size_t allocSize, void*& mem) {
+
+  // try to allocate large pages
+  mem = aligned_ttmem_alloc_large_pages(allocSize);
+  if (mem)
+      sync_cout << "info string Hash table allocation: Windows large pages used." << sync_endl;
+  else
+      sync_cout << "info string Hash table allocation: Windows large pages not used." << sync_endl;
+
+  // fall back to regular, page aligned, allocation if necessary
+  if (!mem)
+      mem = VirtualAlloc(NULL, allocSize, MEM_RESERVE | MEM_COMMIT, PAGE_READWRITE);
+
+  return mem;
+}
+
 #else
 
 void* aligned_ttmem_alloc(size_t allocSize, void*& mem) {
@@ -322,6 +385,28 @@ void* aligned_ttmem_alloc(size_t allocSize, void*& mem) {
 
 #endif
 
+/// aligned_ttmem_free will free the previously allocated ttmem
+#if defined(_WIN64)
+
+void aligned_ttmem_free(void* mem) {
+
+  if (!VirtualFree(mem, 0, MEM_RELEASE))
+  {
+      DWORD err = GetLastError();
+      std::cerr << "Failed to free transposition table. Error code: 0x" <<
+          std::hex << err << std::dec << std::endl;
+      exit(EXIT_FAILURE);
+  }
+}
+
+#else
+
+void aligned_ttmem_free(void *mem) {
+  free(mem);
+}
+
+#endif
+
 
 namespace WinProcGroup {
 

src/misc.h

@@ -34,6 +34,7 @@ const std::string compiler_info();
 void prefetch(void* addr);
 void start_logger(const std::string& fname);
 void* aligned_ttmem_alloc(size_t size, void*& mem);
+void aligned_ttmem_free(void* mem);
 
 void dbg_hit_on(bool b);
 void dbg_hit_on(bool c, bool b);

src/tt.cpp

@@ -63,7 +63,14 @@ void TranspositionTable::resize(size_t mbSize) {
 
   Threads.main()->wait_for_search_finished();
 
-  free(mem);
+  if (mem)
+      aligned_ttmem_free(mem);
+
+  if (!mbSize)
+  {
+      mem = nullptr;
+      return;
+  }
 
   clusterCount = mbSize * 1024 * 1024 / sizeof(Cluster);
   table = static_cast<Cluster*>(aligned_ttmem_alloc(clusterCount * sizeof(Cluster), mem));