C++20

src/Makefile

@@ -54,7 +54,7 @@ ifeq ($(EXE),)
 	EXE=RubiChess
 endif
 
-CXXFLAGS=-std=c++11 -Wall -pedantic -Wextra -Wshadow
+CXXFLAGS=-std=c++2a -Wall -pedantic -Wextra -Wshadow
 CFLAGS=-Wall -pedantic -Wextra -Wshadow -Wno-implicit-function-declaration
 
 

src/Makefile.clang

@@ -25,10 +25,20 @@ CXX="$(CompilerBaseDir)\bin\clang-cl.exe"
 LD="$(CompilerBaseDir)\bin\lld-link.exe"
 PROFDATATOOLEXE="$(CompilerBaseDir)\bin\llvm-profdata.exe"
 !ELSEIF "$(COMP)" == "icx"
-CompilerBaseDir=C:\Program Files (x86)\Intel\oneAPI\compiler\latest\windows
+CompilerBaseDir=C:\Program Files (x86)\Intel\oneAPI\compiler\latest
+!IF EXIST("$(CompilerBaseDir)\windows")				# Path up to Intel oneAPI 2023.1
+CompilerBaseDir=$(CompilerBaseDir)\windows
+!ENDIF
 CXX="$(CompilerBaseDir)\bin\icx-cl.exe"
-LD="$(CompilerBaseDir)\bin-llvm\lld-link.exe"
-PROFDATATOOLEXE="$(CompilerBaseDir)\bin-llvm\llvm-profdata.exe"
+!IF EXISTS("$(CompilerBaseDir)\bin-llvm")			# Path up to Intel oneAPI 2023.1
+CompilerClangDir=$(CompilerBaseDir)\bin-llvm
+!ELSEIF EXISTS("$(CompilerBaseDir)\bin\compiler")	# Path of Intel oneAPI 2024.0
+CompilerClangDir=$(CompilerBaseDir)\bin\compiler
+!ELSE
+!ERROR Folder with Linker tools not found.
+!ENDIF
+LD="$(CompilerClangDir)\lld-link.exe"
+PROFDATATOOLEXE="$(CompilerClangDir)\llvm-profdata.exe"
 !ELSE
 !ERROR Not supported compiler $(COMP)
 !ENDIF
@@ -38,6 +48,11 @@ OBJ=*.obj
 RELDIR=Release-clang
 ZLIBDIR=zlib
 CXXFLAGS=/EHsc /O2 /Oi /Ot /c -flto -fuse-ld=lld
+!IF "$(COMP)" == "clang"
+CPPFLAGS=/std:c++20
+!ELSE
+CPPFLAGS=-Qstd=c++20
+!ENDIF
 LDFLAGS=/OPT:REF /OPT:ICF advapi32.lib
 PROFEXE=RubiChess-WinProf
 NETURL=https://github.com/Matthies/NN/raw/main/
@@ -103,7 +118,7 @@ ARCH=native
 
 # nativeflags
 !IF "$(ARCH)"=="native" && !EXIST($(NATIVEFLAGS)) && "$(VSCMD_ARG_HOST_ARCH)"=="$(VSCMD_ARG_TGT_ARCH)"
-!IF [@$(CXX) /GX -DCPUTEST $(CPUTEST).cpp -o $(CPUTEST) && @$(CPUTEST) > $(NATIVEFLAGS)] != 0
+!IF [$(CXX) $(CPPFLAGS) /GX -DCPUTEST $(CPUTEST).cpp -o $(CPUTEST) && @$(CPUTEST) > $(NATIVEFLAGS)] != 0
 !ERROR Cannot compile and/or run $(CPUTEST) to determine cpu flags on this machine!
 !ENDIF
 !ENDIF
@@ -221,7 +236,7 @@ compile: $(ZLIBDIR)\zlib.lib
 !IFNDEF GITID
 	@FOR /F "tokens=2" %A IN ('cmd /c "($(GIT) show --name-only --abbrev-commit || echo commit unknown) | findstr /r /c:"^commit"" ') DO @(set GITID=%A&& nmake -f Makefile.clang $* /$(MAKEFLAGS) && exit /B)
 !ELSE
-	@$(CXX) $(CXXFLAGS) $(ARCHFLAGS) $(SOURCE) $(PROFILEFLAGS) -D GITVER=\"$(GITVER)\" -D GITID=\"$(GITID)\"
+	@$(CXX) $(CXXFLAGS) $(CPPFLAGS) $(ARCHFLAGS) $(SOURCE) $(PROFILEFLAGS) -D GITVER=\"$(GITVER)\" -D GITID=\"$(GITID)\"
 !IFDEF USEPROFLIB
 	@$(LD) $(LDFLAGS) /OUT:$(PROFEXE).exe $(OBJ) "$(USEPROFLIB)"
 !ELSE

src/RubiChess.cbp

@@ -51,7 +51,7 @@
 		</Build>
 		<Compiler>
 			<Add option="-Wall" />
-			<Add option="-std=c++11" />
+			<Add option="-std=c++20" />
 			<Add option="-fexceptions" />
 			<Add option="-march=znver1" />
 			<Add option="-DUSE_AVX2" />

src/RubiChess.h

@@ -68,6 +68,7 @@
 #ifdef _MSC_VER
 #define _CRT_SECURE_NO_WARNINGS
 #endif
+#include <bit>
 #include <iostream>
 #include <iomanip>
 #include <stdarg.h>
@@ -211,150 +212,28 @@ typedef unsigned int PieceType;
 #define MORETHANONE(x) ((x) & ((x) - 1))
 #define ONEORZERO(x) (!MORETHANONE(x))
 
-template  <typename T> int popcount_legacy(T v)
-{
-    v = v - ((v >> 1) & (T)~(T)0 / 3);
-    v = (v & (T)~(T)0 / 15 * 3) + ((v >> 2) & (T)~(T)0 / 15 * 3);
-    v = (v + (v >> 4)) & (T)~(T)0 / 255 * 15;
-    return  (T)(v * ((T)~(T)0 / 255)) >> (sizeof(T) - 1) * CHAR_BIT;
-}
+//
+// Use C++20 intrinsics
+//
+#define POPCOUNT(x) popcount(x)
+#define POPCOUNT32(x) popcount(x)
+#define GETMSB(i,x) (i = (sizeof(x) * 8 - countl_zero(x) - 1))
+#define GETLSB(i,x) (i = countr_zero(x))
 
-#if defined(_MSC_VER)
-#ifdef USE_BMI1
-#include <immintrin.h>
-#define GETLSB32(i,x) (i =(int) _tzcnt_u32(x))
-#ifdef _M_X64
-#define GETLSB(i,x) (i =(int) _tzcnt_u64(x))
-inline int pullLsb(U64* x) {
-    int i;
-    i = (int)_tzcnt_u64(*x);
-    *x = _blsr_u64(*x);
-    return i;
-}
-#define GETMSB(i,x) (i = 63 ^ (int) _lzcnt_u64(x))
-inline int pullMsb(U64* x) {
-    int i;
-    i = 63 ^ (int)_lzcnt_u64(*x);
-    *x ^= (1ULL << i);
-    return i;
-}
-#else // _M_X64
-#define GETLSB(i,x) (i =(int) ((x) & 0xffffffff ?  _tzcnt_u32((x) & 0xffffffff) : _tzcnt_u32((x) >> 32) + 32))
-inline int pullLsb(U64* x) {
-    int i;
-    i = (int)(*(x) & 0xffffffff ? _tzcnt_u32(*(x) & 0xffffffff) : _tzcnt_u32(*(x) >> 32) + 32);
-    *x ^= (1ULL << i);
-    return i;
-}
-#define GETMSB(i,x) (i = 63 ^ (int) ((x) >> 32 ? _lzcnt_u32((x) >> 32) : _lzcnt_u32((x) & 0xffffffff) + 32))
 inline int pullMsb(U64* x) {
-    int i;
-    i = 63 ^ (int)((*(x) >> 32) ? _lzcnt_u32(*(x) >> 32) : _lzcnt_u32(*(x) & 0xffffffff) + 32);
-    *x ^= (1ULL << i);
+    unsigned int i;
+    GETMSB(i, *x);
+    *x ^= BITSET(i);
     return i;
 }
-#endif
-#else // USE_BMI1
-
-#define GETLSB32(i,x) _BitScanForward((DWORD*)&(i), (x))
-#if defined(_M_X64) || defined(_M_ARM64)
-#define GETLSB(i,x) _BitScanForward64((DWORD*)&(i), (x))
 inline int pullLsb(U64* x) {
-    DWORD i;
-    _BitScanForward64(&i, *x);
+    unsigned int i;
+    GETLSB(i, *x);
     *x &= *x - 1;
     return i;
 }
-#define GETMSB(i,x) _BitScanReverse64((DWORD*)&(i), (x))
-inline int pullMsb(U64* x) {
-    DWORD i;
-    _BitScanReverse64(&i, *x);
-    *x ^= (1ULL << i);
-    return i;
-}
-#else // _M_X64 || _M_ARM64
-#define GETLSB(i,x) ((x) & 0xffffffff ? _BitScanForward((DWORD*)&(i), (x) & 0xffffffff) : (_BitScanForward((DWORD*)&(i), (x) >> 32), i += 32))
-inline int pullLsb(U64* x) {
-    DWORD i;
-    (*(x) & 0xffffffff ? _BitScanForward(&i, *(x) & 0xffffffff) : (_BitScanForward(&i, *(x) >> 32), i += 32));
-    *x ^= (1ULL << i);
-    return i;
-}
-#define GETMSB(i,x) (((x) >> 32) ? (_BitScanReverse((DWORD*)&i, (x) >> 32), (i) += 32) : _BitScanReverse((DWORD*)&i, (x) & 0xffffffff))
-inline int pullMsb(U64* x) {
-    DWORD i;
-    (*(x) >> 32 ? _BitScanReverse(&i, *(x) >> 32) : (_BitScanReverse(&i, *(x) & 0xffffffff), i += 32));
-    *x ^= (1ULL << i);
-    return i;
-}
 
 
-
-#endif
-#endif
-#if !defined(__clang_major__) && (defined(_M_ARM) || defined(_M_ARM64))
-#define POPCOUNT(x) popcount_legacy(x)
-#define POPCOUNT32(x) popcount_legacy(x)
-#else
-#ifdef _M_X64
-#define POPCOUNT(x) (int)(__popcnt64(x))
-#define POPCOUNT32(x) __popcnt(x)
-#else
-#define POPCOUNT(x) (int)(__popcnt((unsigned int)((x) >> 32)) + __popcnt((unsigned int)((x) & 0xffffffff)))
-#define POPCOUNT32(x) __popcnt(x)
-#endif
-#endif
-#else // _MSC_VER
-#ifdef USE_BMI1
-#define GETLSB32(i,x) (i =(int) _tzcnt_u32(x))
-#ifdef IS_64BIT
-#define GETLSB(i,x) (i =  _tzcnt_u64(x))
-inline int pullLsb(U64* x) {
-    int i = _tzcnt_u64(*x);
-    *x = _blsr_u64(*x);
-    return i;
-}
-#define GETMSB(i,x) (i = (63 ^ _lzcnt_u64(x)))
-inline int pullMsb(U64* x) {
-    int i = 63 ^ _lzcnt_u64(*x);
-    *x ^= (1ULL << i);
-    return i;
-}
-#else // IS_64BIT
-#define GETLSB(i,x) (i =(int) ((x) & 0xffffffff ?  _tzcnt_u32((x) & 0xffffffff) : _tzcnt_u32((x) >> 32) + 32))
-inline int pullLsb(U64* x) {
-    int i;
-    i = (int)(*(x) & 0xffffffff ? _tzcnt_u32(*(x) & 0xffffffff) : _tzcnt_u32(*(x) >> 32) + 32);
-    *x ^= (1ULL << i);
-    return i;
-}
-#define GETMSB(i,x) (i = 63 ^ (int) ((x) >> 32 ? _lzcnt_u32((x) >> 32) : _lzcnt_u32((x) & 0xffffffff) + 32))
-inline int pullMsb(U64* x) {
-    int i;
-    i = 63 ^ (int)((*(x) >> 32) ? _lzcnt_u32(*(x) >> 32) : _lzcnt_u32(*(x) & 0xffffffff) + 32);
-    *x ^= (1ULL << i);
-    return i;
-}
-#endif
-#else
-#define GETLSB(i,x) (i = __builtin_ctzll(x))
-#define GETLSB32(i,x) (i =(int) __builtin_ctz(x))
-inline int pullLsb(U64* x) {
-    int i = __builtin_ctzll(*x);
-    *x &= *x - 1;
-    return i;
-}
-#define GETMSB(i,x) (i = (63 ^ __builtin_clzll(x)))
-inline int pullMsb(U64* x) {
-    int i = 63 - __builtin_clzll(*x);
-    *x ^= (1ULL << i);
-    return i;
-}
-#endif
-#define POPCOUNT(x) __builtin_popcountll(x)
-#define POPCOUNT32(x) __builtin_popcount(x)
-#endif
-
 enum Color { WHITE, BLACK };
 #define WHITEBB 0x55aa55aa55aa55aa
 #define BLACKBB 0xaa55aa55aa55aa55

src/RubiChess.vcxproj

@@ -189,6 +189,7 @@
       <WarningLevel>Level3</WarningLevel>
       <Optimization>Disabled</Optimization>
       <PreprocessorDefinitions>_DEBUG;_CONSOLE;%(PreprocessorDefinitions);USE_SSE2;USE_SSSE3;USE_POPCNT;USE_BMI1;USE_AVX2;USE_ZLIB</PreprocessorDefinitions>
+      <LanguageStandard>stdcpp20</LanguageStandard>
     </ClCompile>
     <Link>
       <SubSystem>Console</SubSystem>
@@ -222,6 +223,7 @@
       <WarningLevel>Level3</WarningLevel>
       <Optimization>Disabled</Optimization>
       <PreprocessorDefinitions>_DEBUG;_CONSOLE;%(PreprocessorDefinitions);USE_SSE2;USE_SSSE3;USE_POPCNT;USE_BMI1;USE_AVX2;USE_ZLIB</PreprocessorDefinitions>
+      <LanguageStandard>stdcpp20</LanguageStandard>
     </ClCompile>
     <Link>
       <SubSystem>Console</SubSystem>
@@ -255,6 +257,7 @@
       <WarningLevel>Level3</WarningLevel>
       <Optimization>Disabled</Optimization>
       <PreprocessorDefinitions>_DEBUG;_CONSOLE;%(PreprocessorDefinitions);USE_NEON;USE_ZLIB</PreprocessorDefinitions>
+      <LanguageStandard>stdcpp20</LanguageStandard>
     </ClCompile>
     <Link>
       <SubSystem>Console</SubSystem>
@@ -288,6 +291,7 @@
       <WarningLevel>Level3</WarningLevel>
       <Optimization>Disabled</Optimization>
       <PreprocessorDefinitions>_DEBUG;_CONSOLE;%(PreprocessorDefinitions);USE_NEON;USE_ARM64;USE_ZLIB</PreprocessorDefinitions>
+      <LanguageStandard>stdcpp20</LanguageStandard>
     </ClCompile>
     <Link>
       <SubSystem>Console</SubSystem>
@@ -325,6 +329,7 @@
       <PreprocessorDefinitions>_CONSOLE;%(PreprocessorDefinitions);USE_SSE2;USE_SSSE3;USE_POPCNT;USE_BMI1;USE_AVX2;USE_ZLIB</PreprocessorDefinitions>
       <FavorSizeOrSpeed>Speed</FavorSizeOrSpeed>
       <RuntimeLibrary>MultiThreaded</RuntimeLibrary>
+      <LanguageStandard>stdcpp20</LanguageStandard>
     </ClCompile>
     <Link>
       <SubSystem>Console</SubSystem>
@@ -366,6 +371,7 @@
       <PreprocessorDefinitions>_CONSOLE;%(PreprocessorDefinitions);USE_SSE2;USE_SSSE3;USE_POPCNT;xUSE_BMI1;USE_AVX2;USE_ZLIB</PreprocessorDefinitions>
       <FavorSizeOrSpeed>Speed</FavorSizeOrSpeed>
       <RuntimeLibrary>MultiThreaded</RuntimeLibrary>
+      <LanguageStandard>stdcpp20</LanguageStandard>
     </ClCompile>
     <Link>
       <SubSystem>Console</SubSystem>
@@ -407,6 +413,7 @@
       <PreprocessorDefinitions>_CONSOLE;%(PreprocessorDefinitions);USE_NEON;USE_ZLIB</PreprocessorDefinitions>
       <FavorSizeOrSpeed>Speed</FavorSizeOrSpeed>
       <RuntimeLibrary>MultiThreaded</RuntimeLibrary>
+      <LanguageStandard>stdcpp20</LanguageStandard>
     </ClCompile>
     <Link>
       <SubSystem>Console</SubSystem>
@@ -448,6 +455,7 @@
       <PreprocessorDefinitions>_CONSOLE;%(PreprocessorDefinitions);USE_NEON;USE_ARM64;USE_ZLIB</PreprocessorDefinitions>
       <FavorSizeOrSpeed>Speed</FavorSizeOrSpeed>
       <RuntimeLibrary>MultiThreaded</RuntimeLibrary>
+      <LanguageStandard>stdcpp20</LanguageStandard>
     </ClCompile>
     <Link>
       <SubSystem>Console</SubSystem>

src/nnue.cpp

@@ -672,14 +672,14 @@ typedef uint16x8_t vec128_t;
 #if defined(USE_PROPAGATESPARSE)
 alignas(64) static const array<array<uint16_t, 8>, 256> lookup_indices = []() {
     array<array<uint16_t, 8>, 256> v{};
-    for (int i = 0; i < 256; ++i)
+    for (unsigned int i = 0; i < 256; ++i)
     {
-        int j = i;
+        unsigned int j = i;
         int k = 0;
         while (j)
         {
             unsigned int lsbIndex;
-            GETLSB32(lsbIndex, j);
+            GETLSB(lsbIndex, j);
             j &= j - 1;
             v[i][k++] = lsbIndex;
         }