# 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
+# vnni = 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
# at the end of the line for flag values.
### 2.1. General and architecture defaults
+
+ifeq ($(ARCH),)
+ empty_arch = yes
+endif
+
optimize = yes
debug = no
sanitize = no
avx2 = no
pext = no
avx512 = no
+vnni = no
neon = no
ARCH = x86-64-modern
### 2.2 Architecture specific
+
ifeq ($(ARCH),general-32)
arch = any
bits = 32
ssse3 = yes
endif
-ifeq ($(ARCH),x86-64-modern)
+ifeq ($(ARCH),$(filter $(ARCH),x86-64-sse41-popcnt x86-64-modern))
arch = x86_64
prefetch = yes
popcnt = yes
sse41 = yes
endif
-ifeq ($(ARCH),x86-64-sse41-popcnt)
+ifeq ($(ARCH),x86-64-avx2)
arch = x86_64
prefetch = yes
popcnt = yes
sse = yes
ssse3 = yes
sse41 = yes
+ avx2 = yes
endif
-ifeq ($(ARCH),x86-64-avx2)
+ifeq ($(ARCH),x86-64-bmi2)
arch = x86_64
prefetch = yes
popcnt = yes
ssse3 = yes
sse41 = yes
avx2 = yes
+ pext = yes
endif
-ifeq ($(ARCH),x86-64-bmi2)
+ifeq ($(ARCH),x86-64-avx512)
arch = x86_64
prefetch = yes
popcnt = yes
sse41 = yes
avx2 = yes
pext = yes
+ avx512 = yes
endif
-ifeq ($(ARCH),x86-64-avx512)
+ifeq ($(ARCH),x86-64-vnni)
arch = x86_64
prefetch = yes
popcnt = yes
avx2 = yes
pext = yes
avx512 = yes
+ vnni = yes
endif
ifeq ($(ARCH),armv7)
endif
endif
+ifeq ($(vnni),yes)
+ CXXFLAGS += -DUSE_VNNI
+ ifeq ($(comp),$(filter $(comp),gcc clang mingw))
+ CXXFLAGS += -mavx512vnni -mavx512dq -mavx512vl
+ endif
+endif
+
ifeq ($(sse41),yes)
CXXFLAGS += -DUSE_SSE41
ifeq ($(comp),$(filter $(comp),gcc clang mingw))
endif
ifeq ($(arch),x86_64)
- CXXFLAGS += -DUSE_SSE2
+ CXXFLAGS += -msse2 -DUSE_SSE2
endif
### 3.7 pext
@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-vnni > x86 64-bit with vnni support"
@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 ""
@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 "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 "Advanced examples, for experienced users: "
+ @echo "Advanced examples, for experienced users looking for performance: "
@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 ""
- @echo "The selected architecture $(ARCH) enables the following configuration: "
+ @echo "make help ARCH=x86-64-bmi2"
+ @echo "make -j profile-build ARCH=x86-64-bmi2 COMP=gcc COMPCXX=g++-9.0"
+ @echo "make -j build ARCH=x86-64-ssse3 COMP=clang"
@echo ""
+ifneq ($(empty_arch), yes)
+ @echo "-------------------------------\n"
+ @echo "The selected architecture $(ARCH) will enable the following configuration: "
@$(MAKE) ARCH=$(ARCH) COMP=$(COMP) config-sanity
+endif
.PHONY: help build profile-build strip install clean net objclean profileclean \
build: 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) $(profile_make)
@echo "avx2: '$(avx2)'"
@echo "pext: '$(pext)'"
@echo "avx512: '$(avx512)'"
+ @echo "vnni: '$(vnni)'"
@echo "neon: '$(neon)'"
@echo ""
@echo "Flags:"
@test "$(avx2)" = "yes" || test "$(avx2)" = "no"
@test "$(pext)" = "yes" || test "$(pext)" = "no"
@test "$(avx512)" = "yes" || test "$(avx512)" = "no"
+ @test "$(vnni)" = "yes" || test "$(vnni)" = "no"
@test "$(neon)" = "yes" || test "$(neon)" = "no"
@test "$(comp)" = "gcc" || test "$(comp)" = "icc" || test "$(comp)" = "mingw" || test "$(comp)" = "clang"