X-Git-Url: https://git.sesse.net/?p=stockfish;a=blobdiff_plain;f=src%2FMakefile;h=b021ba1214d8513bccd9550b45956428586599d7;hp=3d84f4826547076ef241e5f07da09c231d63282b;hb=bc654257e74af3ffe837c0fff6e663a073ab8bbf;hpb=21df37d7fd4dcc9b4a9c319382cc43685c0259c8 diff --git a/src/Makefile b/src/Makefile index 3d84f482..b021ba12 100644 --- a/src/Makefile +++ b/src/Makefile @@ -1,7 +1,5 @@ # Stockfish, a UCI chess playing engine derived from Glaurung 2.1 -# Copyright (C) 2004-2008 Tord Romstad (Glaurung author) -# Copyright (C) 2008-2015 Marco Costalba, Joona Kiiski, Tord Romstad -# Copyright (C) 2015-2019 Marco Costalba, Joona Kiiski, Gary Linscott, Tord Romstad +# Copyright (C) 2004-2021 The Stockfish developers (see AUTHORS file) # # Stockfish is free software: you can redistribute it and/or modify # it under the terms of the GNU General Public License as published by @@ -33,13 +31,17 @@ PREFIX = /usr/local BINDIR = $(PREFIX)/bin ### Built-in benchmark for pgo-builds -PGOBENCH = ./$(EXE) bench +ifeq ($(SDE_PATH),) + PGOBENCH = ./$(EXE) bench +else + PGOBENCH = $(SDE_PATH) -- ./$(EXE) bench +endif ### Source and object files SRCS = benchmark.cpp bitbase.cpp bitboard.cpp endgame.cpp evaluate.cpp main.cpp \ material.cpp misc.cpp movegen.cpp movepick.cpp pawns.cpp position.cpp psqt.cpp \ search.cpp thread.cpp timeman.cpp tt.cpp uci.cpp ucioption.cpp tune.cpp syzygy/tbprobe.cpp \ - nnue/evaluate_nnue.cpp nnue/features/half_kp.cpp + nnue/evaluate_nnue.cpp nnue/features/half_ka_v2.cpp OBJS = $(notdir $(SRCS:.cpp=.o)) @@ -59,135 +61,201 @@ endif # ---------------------------------------------------------------------------- # # debug = yes/no --- -DNDEBUG --- Enable/Disable debug mode -# sanitize = undefined/thread/no (-fsanitize ) +# sanitize = none/ ... (-fsanitize ) # --- ( undefined ) --- enable undefined behavior checks -# --- ( thread ) --- enable threading error checks +# --- ( thread ) --- enable threading error checks +# --- ( address ) --- enable memory access checks +# --- ...etc... --- see compiler documentation for supported sanitizers # optimize = yes/no --- (-O3/-fast etc.) --- Enable/Disable optimizations # arch = (name) --- (-arch) --- Target architecture # bits = 64/32 --- -DIS_64BIT --- 64-/32-bit operating system # prefetch = yes/no --- -DUSE_PREFETCH --- Use prefetch asm-instruction # popcnt = yes/no --- -DUSE_POPCNT --- Use popcnt asm-instruction +# pext = yes/no --- -DUSE_PEXT --- Use pext x86_64 asm-instruction # sse = yes/no --- -msse --- Use Intel Streaming SIMD Extensions +# mmx = yes/no --- -mmmx --- Use Intel MMX instructions +# sse2 = yes/no --- -msse2 --- Use Intel Streaming SIMD Extensions 2 # ssse3 = yes/no --- -mssse3 --- Use Intel Supplemental Streaming SIMD Extensions 3 # sse41 = yes/no --- -msse4.1 --- Use Intel Streaming SIMD Extensions 4.1 # avx2 = yes/no --- -mavx2 --- Use Intel Advanced Vector Extensions 2 -# pext = yes/no --- -DUSE_PEXT --- Use pext x86_64 asm-instruction # avx512 = yes/no --- -mavx512bw --- Use Intel Advanced Vector Extensions 512 +# vnni256 = yes/no --- -mavx512vnni --- Use Intel Vector Neural Network Instructions 256 +# vnni512 = yes/no --- -mavx512vnni --- Use Intel Vector Neural Network Instructions 512 # neon = yes/no --- -DUSE_NEON --- Use ARM SIMD architecture # # Note that Makefile is space sensitive, so when adding new architectures # or modifying existing flags, you have to make sure there are no extra spaces # at the end of the line for flag values. +# +# Example of use for these flags: +# make build ARCH=x86-64-avx512 debug=on sanitize="address undefined" + ### 2.1. General and architecture defaults + +ifeq ($(ARCH),) + ARCH = x86-64-modern + help_skip_sanity = yes +endif +# explicitly check for the list of supported architectures (as listed with make help), +# the user can override with `make ARCH=x86-32-vnni256 SUPPORTED_ARCH=true` +ifeq ($(ARCH), $(filter $(ARCH), \ + x86-64-vnni512 x86-64-vnni256 x86-64-avx512 x86-64-bmi2 x86-64-avx2 \ + x86-64-sse41-popcnt x86-64-modern x86-64-ssse3 x86-64-sse3-popcnt \ + x86-64 x86-32-sse41-popcnt x86-32-sse2 x86-32 ppc-64 ppc-32 e2k \ + armv7 armv7-neon armv8 apple-silicon general-64 general-32)) + SUPPORTED_ARCH=true +else + SUPPORTED_ARCH=false +endif + optimize = yes debug = no -sanitize = no +sanitize = none bits = 64 prefetch = no popcnt = no -mmx = no +pext = no sse = no +mmx = no +sse2 = no ssse3 = no sse41 = no avx2 = no -pext = no avx512 = no +vnni256 = no +vnni512 = no neon = no -ARCH = x86-64-modern +STRIP = strip ### 2.2 Architecture specific -ifeq ($(ARCH),general-32) - arch = any - bits = 32 -endif -ifeq ($(ARCH),x86-32-old) - arch = i386 - bits = 32 -endif +ifeq ($(findstring x86,$(ARCH)),x86) -ifeq ($(ARCH),x86-32) +# x86-32/64 + +ifeq ($(findstring x86-32,$(ARCH)),x86-32) arch = i386 bits = 32 - prefetch = yes + sse = yes mmx = yes +else + arch = x86_64 sse = yes + sse2 = yes endif -ifeq ($(ARCH),general-64) - arch = any +ifeq ($(findstring -sse,$(ARCH)),-sse) + sse = yes endif -ifeq ($(ARCH),x86-64) - arch = x86_64 - prefetch = yes +ifeq ($(findstring -popcnt,$(ARCH)),-popcnt) + popcnt = yes +endif + +ifeq ($(findstring -mmx,$(ARCH)),-mmx) + mmx = yes +endif + +ifeq ($(findstring -sse2,$(ARCH)),-sse2) sse = yes + sse2 = yes endif -ifeq ($(ARCH),x86-64-sse3-popcnt) - arch = x86_64 - prefetch = yes +ifeq ($(findstring -ssse3,$(ARCH)),-ssse3) sse = yes - popcnt = yes + sse2 = yes + ssse3 = yes endif -ifeq ($(ARCH),x86-64-ssse3) - arch = x86_64 - prefetch = yes +ifeq ($(findstring -sse41,$(ARCH)),-sse41) sse = yes + sse2 = yes ssse3 = yes + sse41 = yes endif -ifeq ($(ARCH),x86-64-modern) - arch = x86_64 - prefetch = yes +ifeq ($(findstring -modern,$(ARCH)),-modern) popcnt = yes sse = yes + sse2 = yes ssse3 = yes sse41 = yes endif -ifeq ($(ARCH),x86-64-sse41-popcnt) - arch = x86_64 - prefetch = yes +ifeq ($(findstring -avx2,$(ARCH)),-avx2) popcnt = yes sse = yes + sse2 = yes ssse3 = yes sse41 = yes + avx2 = yes endif -ifeq ($(ARCH),x86-64-avx2) - arch = x86_64 - prefetch = yes +ifeq ($(findstring -bmi2,$(ARCH)),-bmi2) popcnt = yes sse = yes + sse2 = yes ssse3 = yes sse41 = yes avx2 = yes + pext = yes endif -ifeq ($(ARCH),x86-64-bmi2) - arch = x86_64 - prefetch = yes +ifeq ($(findstring -avx512,$(ARCH)),-avx512) popcnt = yes sse = yes + sse2 = yes ssse3 = yes sse41 = yes avx2 = yes pext = yes + avx512 = yes endif -ifeq ($(ARCH),x86-64-avx512) - arch = x86_64 - prefetch = yes +ifeq ($(findstring -vnni256,$(ARCH)),-vnni256) + popcnt = yes + sse = yes + sse2 = yes + ssse3 = yes + sse41 = yes + avx2 = yes + pext = yes + vnni256 = yes +endif + +ifeq ($(findstring -vnni512,$(ARCH)),-vnni512) popcnt = yes sse = yes + sse2 = yes ssse3 = yes sse41 = yes avx2 = yes pext = yes avx512 = yes + vnni512 = yes +endif + +ifeq ($(sse),yes) + prefetch = yes +endif + +# 64-bit pext is not available on x86-32 +ifeq ($(bits),32) + pext = no +endif + +else + +# all other architectures + +ifeq ($(ARCH),general-32) + arch = any + bits = 32 +endif + +ifeq ($(ARCH),general-64) + arch = any endif ifeq ($(ARCH),armv7) @@ -196,8 +264,16 @@ ifeq ($(ARCH),armv7) bits = 32 endif +ifeq ($(ARCH),armv7-neon) + arch = armv7 + prefetch = yes + popcnt = yes + neon = yes + bits = 32 +endif + ifeq ($(ARCH),armv8) - arch = armv8-a + arch = armv8 prefetch = yes popcnt = yes neon = yes @@ -221,6 +297,19 @@ ifeq ($(ARCH),ppc-64) prefetch = yes endif +ifeq ($(findstring e2k,$(ARCH)),e2k) + arch = e2k + mmx = yes + bits = 64 + sse = yes + sse2 = yes + ssse3 = yes + sse41 = yes + popcnt = yes +endif + +endif + ### ========================================================================== ### Section 3. Low-level Configuration ### ========================================================================== @@ -239,7 +328,7 @@ ifeq ($(COMP),gcc) CXX=g++ CXXFLAGS += -pedantic -Wextra -Wshadow - ifeq ($(ARCH),$(filter $(ARCH),armv7 armv8)) + ifeq ($(arch),$(filter $(arch),armv7 armv8)) ifeq ($(OS),Android) CXXFLAGS += -m$(bits) LDFLAGS += -m$(bits) @@ -249,12 +338,13 @@ ifeq ($(COMP),gcc) LDFLAGS += -m$(bits) endif + ifeq ($(arch),$(filter $(arch),armv7)) + LDFLAGS += -latomic + endif + ifneq ($(KERNEL),Darwin) LDFLAGS += -Wl,--no-as-needed endif - - gccversion = $(shell $(CXX) --version) - gccisclang = $(findstring clang,$(gccversion)) endif ifeq ($(COMP),mingw) @@ -295,11 +385,15 @@ ifeq ($(COMP),clang) ifneq ($(KERNEL),Darwin) ifneq ($(KERNEL),OpenBSD) + ifneq ($(KERNEL),FreeBSD) + ifneq ($(RTLIB),compiler-rt) LDFLAGS += -latomic endif endif + endif + endif - ifeq ($(ARCH),$(filter $(ARCH),armv7 armv8)) + ifeq ($(arch),$(filter $(arch),armv7 armv8)) ifeq ($(OS),Android) CXXFLAGS += -m$(bits) LDFLAGS += -m$(bits) @@ -310,23 +404,44 @@ ifeq ($(COMP),clang) endif endif +ifeq ($(KERNEL),Darwin) + CXXFLAGS += -mmacosx-version-min=10.14 + LDFLAGS += -mmacosx-version-min=10.14 + ifneq ($(arch),any) + CXXFLAGS += -arch $(arch) + LDFLAGS += -arch $(arch) + endif + XCRUN = xcrun +endif + +# To cross-compile for Android, NDK version r21 or later is recommended. +# In earlier NDK versions, you'll need to pass -fno-addrsig if using GNU binutils. +# Currently we don't know how to make PGO builds with the NDK yet. +ifeq ($(COMP),ndk) + CXXFLAGS += -stdlib=libc++ -fPIE + comp=clang + ifeq ($(arch),armv7) + CXX=armv7a-linux-androideabi16-clang++ + CXXFLAGS += -mthumb -march=armv7-a -mfloat-abi=softfp -mfpu=neon + STRIP=arm-linux-androideabi-strip + endif + ifeq ($(arch),armv8) + CXX=aarch64-linux-android21-clang++ + STRIP=aarch64-linux-android-strip + endif + LDFLAGS += -static-libstdc++ -pie -lm -latomic +endif + ifeq ($(comp),icc) profile_make = icc-profile-make profile_use = icc-profile-use -else -ifeq ($(comp),clang) +else ifeq ($(comp),clang) profile_make = clang-profile-make profile_use = clang-profile-use else profile_make = gcc-profile-make profile_use = gcc-profile-use endif -endif - -ifeq ($(KERNEL),Darwin) - CXXFLAGS += -arch $(arch) -mmacosx-version-min=10.14 - LDFLAGS += -arch $(arch) -mmacosx-version-min=10.14 -endif ### Travis CI script uses COMPILER to overwrite CXX ifdef COMPILER @@ -338,13 +453,26 @@ ifdef COMPCXX CXX=$(COMPCXX) endif +### Sometimes gcc is really clang +ifeq ($(COMP),gcc) + gccversion = $(shell $(CXX) --version) + gccisclang = $(findstring clang,$(gccversion)) + ifneq ($(gccisclang),) + profile_make = clang-profile-make + profile_use = clang-profile-use + endif +endif + ### On mingw use Windows threads, otherwise POSIX ifneq ($(comp),mingw) + CXXFLAGS += -DUSE_PTHREADS # On Android Bionic's C library comes with its own pthread implementation bundled in ifneq ($(OS),Android) # Haiku has pthreads in its libroot, so only link it in on other platforms ifneq ($(KERNEL),Haiku) - LDFLAGS += -lpthread + ifneq ($(COMP),ndk) + LDFLAGS += -lpthread + endif endif endif endif @@ -357,9 +485,9 @@ else endif ### 3.2.2 Debugging with undefined behavior sanitizers -ifneq ($(sanitize),no) - CXXFLAGS += -g3 -fsanitize=$(sanitize) - LDFLAGS += -fsanitize=$(sanitize) +ifneq ($(sanitize),none) + CXXFLAGS += -g3 $(addprefix -fsanitize=,$(sanitize)) + LDFLAGS += $(addprefix -fsanitize=,$(sanitize)) endif ### 3.3 Optimization @@ -378,6 +506,10 @@ ifeq ($(optimize),yes) CXXFLAGS += -mdynamic-no-pic endif endif + + ifeq ($(comp),clang) + CXXFLAGS += -fexperimental-new-pass-manager + endif endif ### 3.4 Bits @@ -389,7 +521,6 @@ endif ifeq ($(prefetch),yes) ifeq ($(sse),yes) CXXFLAGS += -msse - DEPENDFLAGS += -msse endif else CXXFLAGS += -DNO_PREFETCH @@ -397,7 +528,7 @@ endif ### 3.6 popcnt ifeq ($(popcnt),yes) - ifeq ($(arch),$(filter $(arch),ppc64 armv8-a arm64)) + ifeq ($(arch),$(filter $(arch),ppc64 armv7 armv8 arm64)) CXXFLAGS += -DUSE_POPCNT else ifeq ($(comp),icc) CXXFLAGS += -msse3 -DUSE_POPCNT @@ -416,7 +547,21 @@ endif ifeq ($(avx512),yes) CXXFLAGS += -DUSE_AVX512 ifeq ($(comp),$(filter $(comp),gcc clang mingw)) - CXXFLAGS += -mavx512bw + CXXFLAGS += -mavx512f -mavx512bw + endif +endif + +ifeq ($(vnni256),yes) + CXXFLAGS += -DUSE_VNNI + ifeq ($(comp),$(filter $(comp),gcc clang mingw)) + CXXFLAGS += -mavx512f -mavx512bw -mavx512vnni -mavx512dq -mavx512vl -mprefer-vector-width=256 + endif +endif + +ifeq ($(vnni512),yes) + CXXFLAGS += -DUSE_VNNI + ifeq ($(comp),$(filter $(comp),gcc clang mingw)) + CXXFLAGS += -mavx512vnni -mavx512dq -mavx512vl endif endif @@ -434,6 +579,13 @@ ifeq ($(ssse3),yes) endif endif +ifeq ($(sse2),yes) + CXXFLAGS += -DUSE_SSE2 + ifeq ($(comp),$(filter $(comp),gcc clang mingw)) + CXXFLAGS += -msse2 + endif +endif + ifeq ($(mmx),yes) CXXFLAGS += -DUSE_MMX ifeq ($(comp),$(filter $(comp),gcc clang mingw)) @@ -443,10 +595,13 @@ endif ifeq ($(neon),yes) CXXFLAGS += -DUSE_NEON -endif - -ifeq ($(arch),x86_64) - CXXFLAGS += -DUSE_SSE2 + ifeq ($(KERNEL),Linux) + ifneq ($(COMP),ndk) + ifneq ($(arch),armv8) + CXXFLAGS += -mfpu=neon + endif + endif + endif endif ### 3.7 pext @@ -463,7 +618,12 @@ endif ifeq ($(optimize),yes) ifeq ($(debug), no) ifeq ($(comp),clang) - CXXFLAGS += -flto=thin + CXXFLAGS += -flto + ifneq ($(findstring MINGW,$(KERNEL)),) + CXXFLAGS += -fuse-ld=lld + else ifneq ($(findstring MSYS,$(KERNEL)),) + CXXFLAGS += -fuse-ld=lld + endif LDFLAGS += $(CXXFLAGS) # GCC and CLANG use different methods for parallelizing LTO and CLANG pretends to be @@ -472,20 +632,27 @@ ifeq ($(debug), no) ifeq ($(gccisclang),) CXXFLAGS += -flto LDFLAGS += $(CXXFLAGS) -flto=jobserver + ifneq ($(findstring MINGW,$(KERNEL)),) + LDFLAGS += -save-temps + else ifneq ($(findstring MSYS,$(KERNEL)),) + LDFLAGS += -save-temps + endif else - CXXFLAGS += -flto=thin + CXXFLAGS += -flto LDFLAGS += $(CXXFLAGS) endif # To use LTO and static linking on windows, the tool chain requires a recent gcc: -# gcc version 10.1 in msys2 or TDM-GCC version 9.2 are know to work, older might not. +# gcc version 10.1 in msys2 or TDM-GCC version 9.2 are known to work, older might not. # So, only enable it for a cross from Linux by default. else ifeq ($(comp),mingw) ifeq ($(KERNEL),Linux) + ifneq ($(arch),i386) CXXFLAGS += -flto LDFLAGS += $(CXXFLAGS) -flto=jobserver endif endif + endif endif endif @@ -500,6 +667,7 @@ endif ### Section 4. Public Targets ### ========================================================================== + help: @echo "" @echo "To compile stockfish, type: " @@ -508,29 +676,35 @@ help: @echo "" @echo "Supported targets:" @echo "" + @echo "help > Display architecture details" @echo "build > Standard build" - @echo "profile-build > Standard build with PGO" + @echo "net > Download the default nnue net" + @echo "profile-build > Faster build (with profile-guided optimization)" @echo "strip > Strip executable" @echo "install > Install executable" @echo "clean > Clean up" - @echo "net > Download the default nnue net" @echo "" @echo "Supported archs:" @echo "" + @echo "x86-64-vnni512 > x86 64-bit with vnni support 512bit wide" + @echo "x86-64-vnni256 > x86 64-bit with vnni support 256bit wide" @echo "x86-64-avx512 > x86 64-bit with avx512 support" @echo "x86-64-bmi2 > x86 64-bit with bmi2 support" @echo "x86-64-avx2 > x86 64-bit with avx2 support" @echo "x86-64-sse41-popcnt > x86 64-bit with sse41 and popcnt support" - @echo "x86-64-modern > the same as previous (x86-64-sse41-popcnt)" + @echo "x86-64-modern > common modern CPU, currently x86-64-sse41-popcnt" @echo "x86-64-ssse3 > x86 64-bit with ssse3 support" @echo "x86-64-sse3-popcnt > x86 64-bit with sse3 and popcnt support" - @echo "x86-64 > x86 64-bit generic" - @echo "x86-32 > x86 32-bit (also enables MMX and SSE)" - @echo "x86-32-old > x86 32-bit fall back for old hardware" + @echo "x86-64 > x86 64-bit generic (with sse2 support)" + @echo "x86-32-sse41-popcnt > x86 32-bit with sse41 and popcnt support" + @echo "x86-32-sse2 > x86 32-bit with sse2 support" + @echo "x86-32 > x86 32-bit generic (with mmx and sse support)" @echo "ppc-64 > PPC 64-bit" @echo "ppc-32 > PPC 32-bit" @echo "armv7 > ARMv7 32-bit" - @echo "armv8 > ARMv8 64-bit" + @echo "armv7-neon > ARMv7 32-bit with popcnt and neon" + @echo "armv8 > ARMv8 64-bit with popcnt and neon" + @echo "e2k > Elbrus 2000" @echo "apple-silicon > Apple silicon ARM64" @echo "general-64 > unspecified 64-bit" @echo "general-32 > unspecified 32-bit" @@ -541,30 +715,37 @@ help: @echo "mingw > Gnu compiler with MinGW under Windows" @echo "clang > LLVM Clang compiler" @echo "icc > Intel compiler" + @echo "ndk > Google NDK to cross-compile for Android" @echo "" @echo "Simple examples. If you don't know what to do, you likely want to run: " @echo "" - @echo "make -j build ARCH=x86-64 (This is for 64-bit systems)" - @echo "make -j build ARCH=x86-32 (This is for 32-bit systems)" - @echo "" - @echo "Advanced examples, for experienced users: " + @echo "make -j build ARCH=x86-64 (A portable, slow compile for 64-bit systems)" + @echo "make -j build ARCH=x86-32 (A portable, slow compile for 32-bit systems)" @echo "" - @echo "make -j build ARCH=x86-64-modern COMP=clang" - @echo "make -j profile-build ARCH=x86-64-bmi2 COMP=gcc COMPCXX=g++-4.8" + @echo "Advanced examples, for experienced users looking for performance: " @echo "" - @echo "The selected architecture $(ARCH) enables the following configuration: " + @echo "make help ARCH=x86-64-bmi2" + @echo "make -j profile-build ARCH=x86-64-bmi2 COMP=gcc COMPCXX=g++-9.0" + @echo "make -j build ARCH=x86-64-ssse3 COMP=clang" @echo "" + @echo "-------------------------------" +ifeq ($(SUPPORTED_ARCH)$(help_skip_sanity), true) + @echo "The selected architecture $(ARCH) will enable the following configuration: " @$(MAKE) ARCH=$(ARCH) COMP=$(COMP) config-sanity +else + @echo "Specify a supported architecture with the ARCH option for more details" + @echo "" +endif .PHONY: help build profile-build strip install clean net objclean profileclean \ config-sanity icc-profile-use icc-profile-make gcc-profile-use gcc-profile-make \ clang-profile-use clang-profile-make -build: config-sanity +build: net config-sanity $(MAKE) ARCH=$(ARCH) COMP=$(COMP) all -profile-build: config-sanity objclean profileclean +profile-build: net config-sanity objclean profileclean @echo "" @echo "Step 1/4. Building instrumented executable ..." $(MAKE) ARCH=$(ARCH) COMP=$(COMP) $(profile_make) @@ -580,23 +761,40 @@ profile-build: config-sanity objclean profileclean $(MAKE) ARCH=$(ARCH) COMP=$(COMP) profileclean strip: - strip $(EXE) + $(STRIP) $(EXE) install: -mkdir -p -m 755 $(BINDIR) -cp $(EXE) $(BINDIR) -strip $(BINDIR)/$(EXE) -#clean all +# clean all clean: objclean profileclean @rm -f .depend *~ core +# evaluation network (nnue) net: - $(eval nnuenet := $(shell grep EvalFile ucioption.cpp | grep Option | sed 's/.*\(nn-[a-z0-9]\{12\}.nnue\).*/\1/')) + $(eval nnuenet := $(shell grep EvalFileDefaultName evaluate.h | grep define | sed 's/.*\(nn-[a-z0-9]\{12\}.nnue\).*/\1/')) @echo "Default net: $(nnuenet)" $(eval nnuedownloadurl := https://tests.stockfishchess.org/api/nn/$(nnuenet)) - $(eval curl_or_wget := $(shell if hash curl 2>/dev/null; then echo "curl -sL"; elif hash wget 2>/dev/null; then echo "wget -qO-"; fi)) - @if test -f "$(nnuenet)"; then echo "Already available."; else echo "Downloading $(nnuedownloadurl)"; $(curl_or_wget) $(nnuedownloadurl) > $(nnuenet); fi + $(eval curl_or_wget := $(shell if hash curl 2>/dev/null; then echo "curl -skL"; elif hash wget 2>/dev/null; then echo "wget -qO-"; fi)) + @if test -f "$(nnuenet)"; then \ + echo "Already available."; \ + else \ + if [ "x$(curl_or_wget)" = "x" ]; then \ + echo "Automatic download failed: neither curl nor wget is installed. Install one of these tools or download the net manually"; exit 1; \ + else \ + echo "Downloading $(nnuedownloadurl)"; $(curl_or_wget) $(nnuedownloadurl) > $(nnuenet);\ + fi; \ + fi; + $(eval shasum_command := $(shell if hash shasum 2>/dev/null; then echo "shasum -a 256 "; elif hash sha256sum 2>/dev/null; then echo "sha256sum "; fi)) + @if [ "x$(shasum_command)" != "x" ]; then \ + if [ "$(nnuenet)" != "nn-"`$(shasum_command) $(nnuenet) | cut -c1-12`".nnue" ]; then \ + echo "Failed download or $(nnuenet) corrupted, please delete!"; exit 1; \ + fi \ + else \ + echo "shasum / sha256sum not found, skipping net validation"; \ + fi # clean binaries and objects objclean: @@ -605,8 +803,11 @@ objclean: # clean auxiliary profiling files profileclean: @rm -rf profdir - @rm -f bench.txt *.gcda *.gcno ./syzygy/*.gcda ./nnue/*.gcda ./nnue/features/*.gcda + @rm -f bench.txt *.gcda *.gcno ./syzygy/*.gcda ./nnue/*.gcda ./nnue/features/*.gcda *.s @rm -f stockfish.profdata *.profraw + @rm -f stockfish.exe.lto_wrapper_args + @rm -f stockfish.exe.ltrans.out + @rm -f ./-lstdc++.res default: help @@ -617,7 +818,7 @@ default: all: $(EXE) .depend -config-sanity: +config-sanity: net @echo "" @echo "Config:" @echo "debug: '$(debug)'" @@ -629,12 +830,16 @@ config-sanity: @echo "os: '$(OS)'" @echo "prefetch: '$(prefetch)'" @echo "popcnt: '$(popcnt)'" + @echo "pext: '$(pext)'" @echo "sse: '$(sse)'" + @echo "mmx: '$(mmx)'" + @echo "sse2: '$(sse2)'" @echo "ssse3: '$(ssse3)'" @echo "sse41: '$(sse41)'" @echo "avx2: '$(avx2)'" - @echo "pext: '$(pext)'" @echo "avx512: '$(avx512)'" + @echo "vnni256: '$(vnni256)'" + @echo "vnni512: '$(vnni512)'" @echo "neon: '$(neon)'" @echo "" @echo "Flags:" @@ -645,22 +850,27 @@ config-sanity: @echo "Testing config sanity. If this fails, try 'make help' ..." @echo "" @test "$(debug)" = "yes" || test "$(debug)" = "no" - @test "$(sanitize)" = "undefined" || test "$(sanitize)" = "thread" || test "$(sanitize)" = "address" || test "$(sanitize)" = "no" @test "$(optimize)" = "yes" || test "$(optimize)" = "no" + @test "$(SUPPORTED_ARCH)" = "true" @test "$(arch)" = "any" || test "$(arch)" = "x86_64" || test "$(arch)" = "i386" || \ - test "$(arch)" = "ppc64" || test "$(arch)" = "ppc" || \ - test "$(arch)" = "armv7" || test "$(arch)" = "armv8-a" || test "$(arch)" = "arm64" + test "$(arch)" = "ppc64" || test "$(arch)" = "ppc" || test "$(arch)" = "e2k" || \ + test "$(arch)" = "armv7" || test "$(arch)" = "armv8" || test "$(arch)" = "arm64" @test "$(bits)" = "32" || test "$(bits)" = "64" @test "$(prefetch)" = "yes" || test "$(prefetch)" = "no" @test "$(popcnt)" = "yes" || test "$(popcnt)" = "no" + @test "$(pext)" = "yes" || test "$(pext)" = "no" @test "$(sse)" = "yes" || test "$(sse)" = "no" + @test "$(mmx)" = "yes" || test "$(mmx)" = "no" + @test "$(sse2)" = "yes" || test "$(sse2)" = "no" @test "$(ssse3)" = "yes" || test "$(ssse3)" = "no" @test "$(sse41)" = "yes" || test "$(sse41)" = "no" @test "$(avx2)" = "yes" || test "$(avx2)" = "no" - @test "$(pext)" = "yes" || test "$(pext)" = "no" @test "$(avx512)" = "yes" || test "$(avx512)" = "no" + @test "$(vnni256)" = "yes" || test "$(vnni256)" = "no" + @test "$(vnni512)" = "yes" || test "$(vnni512)" = "no" @test "$(neon)" = "yes" || test "$(neon)" = "no" - @test "$(comp)" = "gcc" || test "$(comp)" = "icc" || test "$(comp)" = "mingw" || test "$(comp)" = "clang" + @test "$(comp)" = "gcc" || test "$(comp)" = "icc" || test "$(comp)" = "mingw" || test "$(comp)" = "clang" \ + || test "$(comp)" = "armv7a-linux-androideabi16-clang" || test "$(comp)" = "aarch64-linux-android21-clang" $(EXE): $(OBJS) +$(CXX) -o $@ $(OBJS) $(LDFLAGS) @@ -672,21 +882,22 @@ clang-profile-make: all clang-profile-use: - llvm-profdata merge -output=stockfish.profdata *.profraw + $(XCRUN) llvm-profdata merge -output=stockfish.profdata *.profraw $(MAKE) ARCH=$(ARCH) COMP=$(COMP) \ EXTRACXXFLAGS='-fprofile-instr-use=stockfish.profdata' \ EXTRALDFLAGS='-fprofile-use ' \ all gcc-profile-make: + @mkdir -p profdir $(MAKE) ARCH=$(ARCH) COMP=$(COMP) \ - EXTRACXXFLAGS='-fprofile-generate' \ + EXTRACXXFLAGS='-fprofile-generate=profdir' \ EXTRALDFLAGS='-lgcov' \ all gcc-profile-use: $(MAKE) ARCH=$(ARCH) COMP=$(COMP) \ - EXTRACXXFLAGS='-fprofile-use -fno-peel-loops -fno-tracer' \ + EXTRACXXFLAGS='-fprofile-use=profdir -fno-peel-loops -fno-tracer' \ EXTRALDFLAGS='-lgcov' \ all