PSYSTEM_LOGICAL_PROCESSOR_INFORMATION_EX, PDWORD);
typedef bool(*fun2_t)(USHORT, PGROUP_AFFINITY);
typedef bool(*fun3_t)(HANDLE, CONST GROUP_AFFINITY*, PGROUP_AFFINITY);
+typedef bool(*fun4_t)(USHORT, PGROUP_AFFINITY, USHORT, PUSHORT);
+typedef WORD(*fun5_t)();
}
#endif
#include <sys/mman.h>
#endif
-#if defined(__APPLE__) || defined(__ANDROID__) || defined(__OpenBSD__) || (defined(__GLIBCXX__) && !defined(_GLIBCXX_HAVE_ALIGNED_ALLOC) && !defined(_WIN32))
+#if defined(__APPLE__) || defined(__ANDROID__) || defined(__OpenBSD__) || (defined(__GLIBCXX__) && !defined(_GLIBCXX_HAVE_ALIGNED_ALLOC) && !defined(_WIN32)) || defined(__e2k__)
#define POSIXALIGNEDALLOC
#include <stdlib.h>
#endif
using namespace std;
+namespace Stockfish {
+
namespace {
/// Version number. If Version is left empty, then compile date in the format
static Logger l;
- if (!fname.empty() && !l.file.is_open())
+ if (l.file.is_open())
+ {
+ cout.rdbuf(l.out.buf);
+ cin.rdbuf(l.in.buf);
+ l.file.close();
+ }
+
+ if (!fname.empty())
{
l.file.open(fname, ifstream::out);
cin.rdbuf(&l.in);
cout.rdbuf(&l.out);
}
- else if (fname.empty() && l.file.is_open())
- {
- cout.rdbuf(l.out.buf);
- cin.rdbuf(l.in.buf);
- l.file.close();
- }
}
};
compiler += "(version ";
compiler += stringify(_MSC_FULL_VER) "." stringify(_MSC_BUILD);
compiler += ")";
+ #elif defined(__e2k__) && defined(__LCC__)
+ #define dot_ver2(n) \
+ compiler += (char)'.'; \
+ compiler += (char)('0' + (n) / 10); \
+ compiler += (char)('0' + (n) % 10);
+
+ compiler += "MCST LCC ";
+ compiler += "(version ";
+ compiler += std::to_string(__LCC__ / 100);
+ dot_ver2(__LCC__ % 100)
+ dot_ver2(__LCC_MINOR__)
+ compiler += ")";
#elif __GNUC__
compiler += "g++ (GNUC) ";
compiler += make_version_string(__GNUC__, __GNUC_MINOR__, __GNUC_PATCHLEVEL__);
#if defined(_WIN32)
-static void* aligned_large_pages_alloc_win(size_t allocSize) {
+static void* aligned_large_pages_alloc_windows(size_t allocSize) {
+
+ #if !defined(_WIN64)
+ (void)allocSize; // suppress unused-parameter compiler warning
+ return nullptr;
+ #else
HANDLE hProcessToken { };
LUID luid { };
CloseHandle(hProcessToken);
return mem;
+
+ #endif
}
void* aligned_large_pages_alloc(size_t allocSize) {
// Try to allocate large pages
- void* mem = aligned_large_pages_alloc_win(allocSize);
+ void* mem = aligned_large_pages_alloc_windows(allocSize);
// Fall back to regular, page aligned, allocation if necessary
if (!mem)
if (mem && !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;
+ std::cerr << "Failed to free large page memory. Error code: 0x"
+ << std::hex << err
+ << std::dec << std::endl;
exit(EXIT_FAILURE);
}
}
#else
-/// best_group() retrieves logical processor information using Windows specific
-/// API and returns the best group id for the thread with index idx. Original
+/// best_node() retrieves logical processor information using Windows specific
+/// API and returns the best node id for the thread with index idx. Original
/// code from Texel by Peter Ă–sterlund.
-int best_group(size_t idx) {
+int best_node(size_t idx) {
int threads = 0;
int nodes = 0;
if (!fun1)
return -1;
- // First call to get returnLength. We expect it to fail due to null buffer
+ // First call to GetLogicalProcessorInformationEx() to get returnLength.
+ // We expect the call to fail due to null buffer.
if (fun1(RelationAll, nullptr, &returnLength))
return -1;
SYSTEM_LOGICAL_PROCESSOR_INFORMATION_EX *buffer, *ptr;
ptr = buffer = (SYSTEM_LOGICAL_PROCESSOR_INFORMATION_EX*)malloc(returnLength);
- // Second call, now we expect to succeed
+ // Second call to GetLogicalProcessorInformationEx(), now we expect to succeed
if (!fun1(RelationAll, buffer, &returnLength))
{
free(buffer);
void bindThisThread(size_t idx) {
// Use only local variables to be thread-safe
- int group = best_group(idx);
+ int node = best_node(idx);
- if (group == -1)
+ if (node == -1)
return;
// Early exit if the needed API are not available at runtime
HMODULE k32 = GetModuleHandle("Kernel32.dll");
auto fun2 = (fun2_t)(void(*)())GetProcAddress(k32, "GetNumaNodeProcessorMaskEx");
auto fun3 = (fun3_t)(void(*)())GetProcAddress(k32, "SetThreadGroupAffinity");
+ auto fun4 = (fun4_t)(void(*)())GetProcAddress(k32, "GetNumaNodeProcessorMask2");
+ auto fun5 = (fun5_t)(void(*)())GetProcAddress(k32, "GetMaximumProcessorGroupCount");
if (!fun2 || !fun3)
return;
- GROUP_AFFINITY affinity;
- if (fun2(group, &affinity))
- fun3(GetCurrentThread(), &affinity, nullptr);
+ if (!fun4 || !fun5)
+ {
+ GROUP_AFFINITY affinity;
+ if (fun2(node, &affinity)) // GetNumaNodeProcessorMaskEx
+ fun3(GetCurrentThread(), &affinity, nullptr); // SetThreadGroupAffinity
+ }
+ else
+ {
+ // If a numa node has more than one processor group, we assume they are
+ // sized equal and we spread threads evenly across the groups.
+ USHORT elements, returnedElements;
+ elements = fun5(); // GetMaximumProcessorGroupCount
+ GROUP_AFFINITY *affinity = (GROUP_AFFINITY*)malloc(elements * sizeof(GROUP_AFFINITY));
+ if (fun4(node, affinity, elements, &returnedElements)) // GetNumaNodeProcessorMask2
+ fun3(GetCurrentThread(), &affinity[idx % returnedElements], nullptr); // SetThreadGroupAffinity
+ free(affinity);
+ }
}
#endif
} // namespace CommandLine
+
+} // namespace Stockfish