X-Git-Url: https://git.sesse.net/?a=blobdiff_plain;f=src%2FMakefile;h=0b22fb4e6e93eaf8c1b25653ba4bff8659a79c76;hb=66bf45b99e2061c1ba74f9975bc5059ac0121dfd;hp=eff8bacae0abec256c9404bec30c8ca12b2839dc;hpb=e178a09c47e340e5183d9bd2d331741aa837ba8a;p=stockfish diff --git a/src/Makefile b/src/Makefile index eff8baca..0b22fb4e 100644 --- a/src/Makefile +++ b/src/Makefile @@ -1,5 +1,5 @@ # Stockfish, a UCI chess playing engine derived from Glaurung 2.1 -# Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file) +# Copyright (C) 2004-2023 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 @@ -69,32 +69,33 @@ VPATH = syzygy:nnue:nnue/features ### Section 2. High-level Configuration ### ========================================================================== # -# flag --- Comp switch --- Description +# flag --- Comp switch --- Description # ---------------------------------------------------------------------------- # -# debug = yes/no --- -DNDEBUG --- Enable/Disable debug mode +# debug = yes/no --- -DNDEBUG --- Enable/Disable debug mode # sanitize = none/ ... (-fsanitize ) -# --- ( undefined ) --- enable undefined behavior 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 -# avxvnni = yes/no --- -mavxvnni --- Use Intel Vector Neural Network Instructions AVX -# 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 +# --- ( undefined ) --- enable undefined behavior 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 +# avxvnni = yes/no --- -mavxvnni --- Use Intel Vector Neural Network Instructions AVX +# avx512 = yes/no --- -mavx512bw --- Use Intel Advanced Vector Extensions 512 +# vnni256 = yes/no --- -mavx256vnni --- Use Intel Vector Neural Network Instructions 512 with 256bit operands +# vnni512 = yes/no --- -mavx512vnni --- Use Intel Vector Neural Network Instructions 512 +# neon = yes/no --- -DUSE_NEON --- Use ARM SIMD architecture +# dotprod = yes/no --- -DUSE_NEON_DOTPROD --- Use ARM advanced SIMD Int8 dot product instructions # # 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 @@ -116,7 +117,7 @@ ifeq ($(ARCH), $(filter $(ARCH), \ x86-64-vnni512 x86-64-vnni256 x86-64-avx512 x86-64-avxvnni x86-64-bmi2 \ x86-64-avx2 x86-64-sse41-popcnt x86-64-modern x86-64-ssse3 x86-64-sse3-popcnt \ x86-64 x86-32-sse41-popcnt x86-32-sse2 x86-32 ppc-64 ppc-32 e2k \ - armv7 armv7-neon armv8 apple-silicon general-64 general-32)) + armv7 armv7-neon armv8 armv8-dotprod apple-silicon general-64 general-32 riscv64)) SUPPORTED_ARCH=true else SUPPORTED_ARCH=false @@ -140,6 +141,7 @@ avx512 = no vnni256 = no vnni512 = no neon = no +dotprod = no arm_version = 0 STRIP = strip @@ -308,11 +310,21 @@ ifeq ($(ARCH),armv8) arm_version = 8 endif +ifeq ($(ARCH),armv8-dotprod) + arch = armv8 + prefetch = yes + popcnt = yes + neon = yes + dotprod = yes + arm_version = 8 +endif + ifeq ($(ARCH),apple-silicon) arch = arm64 prefetch = yes popcnt = yes neon = yes + dotprod = yes arm_version = 8 endif @@ -338,16 +350,26 @@ ifeq ($(findstring e2k,$(ARCH)),e2k) popcnt = yes endif +ifeq ($(ARCH),riscv64) + arch = riscv64 +endif endif + ### ========================================================================== ### Section 3. Low-level Configuration ### ========================================================================== ### 3.1 Selecting compiler (default = gcc) -CXXFLAGS += -Wall -Wcast-qual -fno-exceptions -std=c++17 $(EXTRACXXFLAGS) -DEPENDFLAGS += -std=c++17 -LDFLAGS += $(EXTRALDFLAGS) +ifeq ($(MAKELEVEL),0) + export ENV_CXXFLAGS := $(CXXFLAGS) + export ENV_DEPENDFLAGS := $(DEPENDFLAGS) + export ENV_LDFLAGS := $(LDFLAGS) +endif + +CXXFLAGS = $(ENV_CXXFLAGS) -Wall -Wcast-qual -fno-exceptions -std=c++17 $(EXTRACXXFLAGS) +DEPENDFLAGS = $(ENV_DEPENDFLAGS) -std=c++17 +LDFLAGS = $(ENV_LDFLAGS) $(EXTRALDFLAGS) ifeq ($(COMP),) COMP=gcc @@ -356,13 +378,16 @@ endif ifeq ($(COMP),gcc) comp=gcc CXX=g++ - CXXFLAGS += -pedantic -Wextra -Wshadow + CXXFLAGS += -pedantic -Wextra -Wshadow -Wmissing-declarations - ifeq ($(arch),$(filter $(arch),armv7 armv8)) + ifeq ($(arch),$(filter $(arch),armv7 armv8 riscv64)) ifeq ($(OS),Android) CXXFLAGS += -m$(bits) LDFLAGS += -m$(bits) endif + ifeq ($(ARCH),riscv64) + CXXFLAGS += -latomic + endif else CXXFLAGS += -m$(bits) LDFLAGS += -m$(bits) @@ -397,7 +422,7 @@ ifeq ($(COMP),mingw) CXX=i686-w64-mingw32-c++-posix endif endif - CXXFLAGS += -pedantic -Wextra -Wshadow + CXXFLAGS += -pedantic -Wextra -Wshadow -Wmissing-declarations endif ifeq ($(COMP),icc) @@ -413,7 +438,8 @@ ifeq ($(COMP),clang) CXX=x86_64-w64-mingw32-clang++ endif - CXXFLAGS += -pedantic -Wextra -Wshadow + CXXFLAGS += -pedantic -Wextra -Wshadow -Wmissing-prototypes \ + -Wconditional-uninitialized ifeq ($(filter $(KERNEL),Darwin OpenBSD FreeBSD),) ifeq ($(target_windows),) @@ -423,11 +449,14 @@ ifeq ($(COMP),clang) endif endif - ifeq ($(arch),$(filter $(arch),armv7 armv8)) + ifeq ($(arch),$(filter $(arch),armv7 armv8 riscv64)) ifeq ($(OS),Android) CXXFLAGS += -m$(bits) LDFLAGS += -m$(bits) endif + ifeq ($(ARCH),riscv64) + CXXFLAGS += -latomic + endif else CXXFLAGS += -m$(bits) LDFLAGS += -m$(bits) @@ -496,7 +525,7 @@ endif ### Sometimes gcc is really clang ifeq ($(COMP),gcc) - gccversion = $(shell $(CXX) --version) + gccversion = $(shell $(CXX) --version 2>/dev/null) gccisclang = $(findstring clang,$(gccversion)) ifneq ($(gccisclang),) profile_make = clang-profile-make @@ -542,20 +571,23 @@ ifeq ($(optimize),yes) endif endif - ifeq ($(KERNEL),Darwin) - ifeq ($(comp),$(filter $(comp),clang icc)) - CXXFLAGS += -mdynamic-no-pic - endif + ifeq ($(KERNEL),Darwin) + ifeq ($(comp),$(filter $(comp),clang icc)) + CXXFLAGS += -mdynamic-no-pic + endif - ifeq ($(comp),gcc) - ifneq ($(arch),arm64) - CXXFLAGS += -mdynamic-no-pic - endif - endif - endif + ifeq ($(comp),gcc) + ifneq ($(arch),arm64) + CXXFLAGS += -mdynamic-no-pic + endif + endif + endif ifeq ($(comp),clang) - CXXFLAGS += -fexperimental-new-pass-manager + clangmajorversion = $(shell $(CXX) -dumpversion 2>/dev/null | cut -f1 -d.) + ifeq ($(shell expr $(clangmajorversion) \< 16),1) + CXXFLAGS += -fexperimental-new-pass-manager + endif endif endif @@ -587,7 +619,7 @@ endif ifeq ($(avx2),yes) CXXFLAGS += -DUSE_AVX2 ifeq ($(comp),$(filter $(comp),gcc clang mingw)) - CXXFLAGS += -mavx2 + CXXFLAGS += -mavx2 -mbmi endif endif @@ -658,6 +690,10 @@ ifeq ($(neon),yes) endif endif +ifeq ($(dotprod),yes) + CXXFLAGS += -march=armv8.2-a+dotprod -DUSE_NEON_DOTPROD +endif + ### 3.7 pext ifeq ($(pext),yes) CXXFLAGS += -DUSE_PEXT @@ -666,13 +702,25 @@ ifeq ($(pext),yes) endif endif +### 3.7.1 Try to include git commit sha for versioning +GIT_SHA = $(shell git rev-parse --short HEAD 2>/dev/null) +ifneq ($(GIT_SHA), ) + CXXFLAGS += -DGIT_SHA=$(GIT_SHA) +endif + +### 3.7.2 Try to include git commit date for versioning +GIT_DATE = $(shell git show -s --date=format:'%Y%m%d' --format=%cd HEAD 2>/dev/null) +ifneq ($(GIT_DATE), ) + CXXFLAGS += -DGIT_DATE=$(GIT_DATE) +endif + ### 3.8 Link Time Optimization ### This is a mix of compile and link time options because the lto link phase ### needs access to the optimization flags. ifeq ($(optimize),yes) ifeq ($(debug), no) ifeq ($(comp),clang) - CXXFLAGS += -flto + CXXFLAGS += -flto=full ifeq ($(target_windows),yes) CXXFLAGS += -fuse-ld=lld endif @@ -682,21 +730,19 @@ ifeq ($(debug), no) # GCC on some systems. else ifeq ($(comp),gcc) ifeq ($(gccisclang),) - CXXFLAGS += -flto + CXXFLAGS += -flto -flto-partition=one LDFLAGS += $(CXXFLAGS) -flto=jobserver else - CXXFLAGS += -flto + CXXFLAGS += -flto=full LDFLAGS += $(CXXFLAGS) endif # To use LTO and static linking on Windows, # the tool chain requires gcc version 10.1 or later. else ifeq ($(comp),mingw) - ifneq ($(arch),i386) - CXXFLAGS += -flto + CXXFLAGS += -flto -flto-partition=one LDFLAGS += $(CXXFLAGS) -save-temps endif - endif endif endif @@ -721,19 +767,19 @@ help: @echo "Supported targets:" @echo "" @echo "help > Display architecture details" - @echo "build > Standard build" + @echo "profile-build > standard build with profile-guided optimization" + @echo "build > skip profile-guided optimization" @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 "" @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-vnni512 > x86 64-bit with vnni 512bit support" + @echo "x86-64-vnni256 > x86 64-bit with vnni 512bit support, limit operands to 256bit wide" @echo "x86-64-avx512 > x86 64-bit with avx512 support" - @echo "x86-64-avxvnni > x86 64-bit with avxvnni support" + @echo "x86-64-avxvnni > x86 64-bit with vnni 256bit 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" @@ -749,10 +795,12 @@ help: @echo "armv7 > ARMv7 32-bit" @echo "armv7-neon > ARMv7 32-bit with popcnt and neon" @echo "armv8 > ARMv8 64-bit with popcnt and neon" + @echo "armv8-dotprod > ARMv8 64-bit with popcnt, neon and dot product support" @echo "e2k > Elbrus 2000" @echo "apple-silicon > Apple silicon ARM64" @echo "general-64 > unspecified 64-bit" @echo "general-32 > unspecified 32-bit" + @echo "riscv64 > RISC-V 64-bit" @echo "" @echo "Supported compilers:" @echo "" @@ -762,14 +810,15 @@ help: @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 "Simple examples. If you don't know what to do, you likely want to run one of: " @echo "" - @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 "make -j profile-build ARCH=x86-64-avx2 # typically a fast compile for common systems " + @echo "make -j profile-build ARCH=x86-64-modern # A more portable compile for 64-bit systems " + @echo "make -j profile-build ARCH=x86-64 # A portable compile for 64-bit systems " @echo "" - @echo "Advanced examples, for experienced users looking for performance: " + @echo "Advanced examples, for experienced users: " @echo "" - @echo "make help ARCH=x86-64-bmi2" + @echo "make -j profile-build 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 "" @@ -785,7 +834,7 @@ 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 + clang-profile-use clang-profile-make FORCE build: net config-sanity $(MAKE) ARCH=$(ARCH) COMP=$(COMP) all @@ -821,25 +870,37 @@ clean: objclean profileclean net: $(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)) + $(eval nnuedownloadurl1 := https://tests.stockfishchess.org/api/nn/$(nnuenet)) + $(eval nnuedownloadurl2 := https://github.com/official-stockfish/networks/raw/master/$(nnuenet)) $(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) > $(nnuenet);\ - fi; \ - fi; + @if [ "x$(curl_or_wget)" = "x" ]; then \ + echo "Neither curl nor wget is installed. Install one of these tools unless the net has been downloaded manually"; \ + 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 $(nnuenet) corrupted, please delete!"; exit 1; \ - fi \ - else \ + @if [ "x$(shasum_command)" = "x" ]; then \ echo "shasum / sha256sum not found, skipping net validation"; \ fi + @for nnuedownloadurl in "$(nnuedownloadurl1)" "$(nnuedownloadurl2)"; do \ + if test -f "$(nnuenet)"; then \ + echo "$(nnuenet) available."; \ + else \ + if [ "x$(curl_or_wget)" != "x" ]; then \ + echo "Downloading $${nnuedownloadurl}"; $(curl_or_wget) $${nnuedownloadurl} > $(nnuenet);\ + else \ + echo "No net found and download not possible"; exit 1;\ + fi; \ + fi; \ + if [ "x$(shasum_command)" != "x" ]; then \ + if [ "$(nnuenet)" != "nn-"`$(shasum_command) $(nnuenet) | cut -c1-12`".nnue" ]; then \ + echo "Removing failed download"; rm -f $(nnuenet); \ + else \ + echo "Network validated"; break; \ + fi; \ + fi; \ + done + @if ! test -f "$(nnuenet)"; then \ + echo "Failed to download $(nnuenet)."; \ + fi # clean binaries and objects objclean: @@ -902,7 +963,7 @@ config-sanity: net @test "$(SUPPORTED_ARCH)" = "true" @test "$(arch)" = "any" || test "$(arch)" = "x86_64" || test "$(arch)" = "i386" || \ test "$(arch)" = "ppc64" || test "$(arch)" = "ppc" || test "$(arch)" = "e2k" || \ - test "$(arch)" = "armv7" || test "$(arch)" = "armv8" || test "$(arch)" = "arm64" + test "$(arch)" = "armv7" || test "$(arch)" = "armv8" || test "$(arch)" = "arm64" || test "$(arch)" = "riscv64" @test "$(bits)" = "32" || test "$(bits)" = "64" @test "$(prefetch)" = "yes" || test "$(prefetch)" = "no" @test "$(popcnt)" = "yes" || test "$(popcnt)" = "no" @@ -923,6 +984,10 @@ config-sanity: net $(EXE): $(OBJS) +$(CXX) -o $@ $(OBJS) $(LDFLAGS) +# Force recompilation to ensure version info is up-to-date +misc.o: FORCE +FORCE: + clang-profile-make: $(MAKE) ARCH=$(ARCH) COMP=$(COMP) \ EXTRACXXFLAGS='-fprofile-instr-generate ' \ @@ -965,4 +1030,6 @@ icc-profile-use: .depend: $(SRCS) -@$(CXX) $(DEPENDFLAGS) -MM $(SRCS) > $@ 2> /dev/null +ifeq (, $(filter $(MAKECMDGOALS), help strip install clean net objclean profileclean config-sanity)) -include .depend +endif