Support VNNI on 256bit vectors

[stockfish] / src / misc.cpp

diff --git a/src/misc.cpp b/src/misc.cpp
index fc3746cf69a112a24407d46fb0488cc8df1fe2d5..80c436ac0f256ad31d8cc935a55498820d1bbdc5 100644 (file)
--- a/src/misc.cpp
+++ b/src/misc.cpp
@@ -51,7 +51,7 @@ typedef bool(*fun3_t)(HANDLE, CONST GROUP_AFFINITY*, PGROUP_AFFINITY);
 #include <sys/mman.h>
 #endif
 
-#if (defined(__APPLE__) && defined(_LIBCPP_HAS_C11_FEATURES)) || 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))
 #define POSIXALIGNEDALLOC
 #include <stdlib.h>
 #endif
@@ -219,9 +219,13 @@ const std::string compiler_info() {
 
   compiler += "\nCompilation settings include: ";
   compiler += (Is64Bit ? " 64bit" : " 32bit");
+  #if defined(USE_VNNI)
+    compiler += " VNNI";
+  #endif
   #if defined(USE_AVX512)
     compiler += " AVX512";
   #endif
+  compiler += (HasPext ? " BMI2" : "");
   #if defined(USE_AVX2)
     compiler += " AVX2";
   #endif
@@ -231,11 +235,17 @@ const std::string compiler_info() {
   #if defined(USE_SSSE3)
     compiler += " SSSE3";
   #endif
-    compiler += (HasPext ? " BMI2" : "");
-    compiler += (HasPopCnt ? " POPCNT" : "");
+  #if defined(USE_SSE2)
+    compiler += " SSE2";
+  #endif
+  compiler += (HasPopCnt ? " POPCNT" : "");
   #if defined(USE_MMX)
     compiler += " MMX";
   #endif
+  #if defined(USE_NEON)
+    compiler += " NEON";
+  #endif
+
   #if !defined(NDEBUG)
     compiler += " DEBUG";
   #endif
@@ -318,17 +328,17 @@ void prefetch(void* addr) {
 
 #endif
 
-/// Wrappers for systems where the c++17 implementation doesn't guarantee the availability of aligned_alloc.
-/// Memory allocated with std_aligned_alloc must be freed with std_aligned_free.
-///
+
+/// std_aligned_alloc() is our wrapper for systems where the c++17 implementation
+/// does not guarantee the availability of aligned_alloc(). Memory allocated with
+/// std_aligned_alloc() must be freed with std_aligned_free().
 
 void* std_aligned_alloc(size_t alignment, size_t size) {
+
 #if defined(POSIXALIGNEDALLOC)
-  void *pointer;
-  if(posix_memalign(&pointer, alignment, size) == 0)
-      return pointer;
-  return nullptr;
-#elif (defined(_WIN32) || (defined(__APPLE__) && !defined(_LIBCPP_HAS_C11_FEATURES)))
+  void *mem;
+  return posix_memalign(&mem, alignment, size) ? nullptr : mem;
+#elif defined(_WIN32)
   return _mm_malloc(size, alignment);
 #else
   return std::aligned_alloc(alignment, size);
@@ -336,16 +346,17 @@ void* std_aligned_alloc(size_t alignment, size_t size) {
 }
 
 void std_aligned_free(void* ptr) {
+
 #if defined(POSIXALIGNEDALLOC)
   free(ptr);
-#elif (defined(_WIN32) || (defined(__APPLE__) && !defined(_LIBCPP_HAS_C11_FEATURES)))
+#elif defined(_WIN32)
   _mm_free(ptr);
 #else
   free(ptr);
 #endif
 }
 
-/// aligned_ttmem_alloc() will return suitably aligned memory, and if possible use large pages.
+/// aligned_ttmem_alloc() will return suitably aligned memory, if possible using large pages.
 /// The returned pointer is the aligned one, while the mem argument is the one that needs
 /// to be passed to free. With c++17 some of this functionality could be simplified.
 
@@ -357,7 +368,9 @@ void* aligned_ttmem_alloc(size_t allocSize, void*& mem) {
   size_t size = ((allocSize + alignment - 1) / alignment) * alignment; // multiple of alignment
   if (posix_memalign(&mem, alignment, size))
      mem = nullptr;
+#if defined(MADV_HUGEPAGE)
   madvise(mem, allocSize, MADV_HUGEPAGE);
+#endif
   return mem;
 }