X-Git-Url: https://git.sesse.net/?p=stockfish;a=blobdiff_plain;f=src%2FMakefile;h=0804cdd51a9dcfe7bafc5feb15b11a4ed0455c4f;hp=b7585a174c6bc87163e584719eeeb339345cd5f4;hb=dd63b98fb06e050aa961fbad6fd1f9316f2b17df;hpb=d7a26899a973536ab9d3ce4771d8276d1a4dc55c

diff --git a/src/Makefile b/src/Makefile
index b7585a17..0804cdd5 100644
--- a/src/Makefile
+++ b/src/Makefile
@@ -68,13 +68,12 @@ endif
 # prefetch = yes/no   --- -DUSE_PREFETCH   --- Use prefetch asm-instruction
 # popcnt = yes/no     --- -DUSE_POPCNT     --- Use popcnt asm-instruction
 # sse = yes/no        --- -msse            --- Use Intel Streaming SIMD Extensions
-# sse3 = yes/no       --- -msse3           --- Use Intel Streaming SIMD Extensions 3
 # ssse3 = yes/no      --- -mssse3          --- Use Intel Supplemental Streaming SIMD Extensions 3
 # sse41 = yes/no      --- -msse4.1         --- Use Intel Streaming SIMD Extensions 4.1
-# sse42 = yes/no      --- -msse4.2         --- Use Intel Streaming SIMD Extensions 4.2
 # 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
@@ -88,14 +87,14 @@ sanitize = no
 bits = 64
 prefetch = no
 popcnt = no
+mmx = no
 sse = no
-sse3 = no
 ssse3 = no
 sse41 = no
-sse42 = no
 avx2 = no
 pext = no
 avx512 = no
+vnni = no
 neon = no
 ARCH = x86-64-modern
 
@@ -114,6 +113,7 @@ ifeq ($(ARCH),x86-32)
 	arch = i386
 	bits = 32
 	prefetch = yes
+	mmx = yes
 	sse = yes
 endif
 
@@ -127,18 +127,10 @@ ifeq ($(ARCH),x86-64)
 	sse = yes
 endif
 
-ifeq ($(ARCH),x86-64-sse3)
-	arch = x86_64
-	prefetch = yes
-	sse = yes
-	sse3 = yes
-endif
-
 ifeq ($(ARCH),x86-64-sse3-popcnt)
 	arch = x86_64
 	prefetch = yes
 	sse = yes
-	sse3 = yes
 	popcnt = yes
 endif
 
@@ -146,78 +138,71 @@ ifeq ($(ARCH),x86-64-ssse3)
 	arch = x86_64
 	prefetch = yes
 	sse = yes
-	sse3 = yes
 	ssse3 = yes
 endif
 
-ifeq ($(ARCH),x86-64-sse41)
+ifeq ($(ARCH),x86-64-modern)
 	arch = x86_64
 	prefetch = yes
 	popcnt = yes
 	sse = yes
-	sse3 = yes
 	ssse3 = yes
 	sse41 = yes
 endif
 
-ifeq ($(ARCH),x86-64-modern)
+ifeq ($(ARCH),x86-64-sse41-popcnt)
 	arch = x86_64
 	prefetch = yes
 	popcnt = yes
 	sse = yes
-	sse3 = yes
 	ssse3 = yes
 	sse41 = yes
 endif
 
-ifeq ($(ARCH),x86-64-sse42)
+ifeq ($(ARCH),x86-64-avx2)
 	arch = x86_64
 	prefetch = yes
 	popcnt = yes
 	sse = yes
-	sse3 = yes
 	ssse3 = yes
 	sse41 = yes
-	sse42 = yes
+	avx2 = yes
 endif
 
-ifeq ($(ARCH),x86-64-avx2)
+ifeq ($(ARCH),x86-64-bmi2)
 	arch = x86_64
 	prefetch = yes
 	popcnt = yes
 	sse = yes
-	sse3 = yes
 	ssse3 = yes
 	sse41 = yes
-	sse42 = yes
 	avx2 = yes
+	pext = yes
 endif
 
-ifeq ($(ARCH),x86-64-bmi2)
+ifeq ($(ARCH),x86-64-avx512)
 	arch = x86_64
 	prefetch = yes
 	popcnt = yes
 	sse = yes
-	sse3 = yes
 	ssse3 = yes
 	sse41 = yes
-	sse42 = 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
 	sse = yes
-	sse3 = yes
 	ssse3 = yes
 	sse41 = yes
-	sse42 = yes
 	avx2 = yes
 	pext = yes
 	avx512 = yes
+	vnni = yes
 endif
 
 ifeq ($(ARCH),armv7)
@@ -282,7 +267,7 @@ ifeq ($(COMP),gcc)
 	ifneq ($(KERNEL),Darwin)
 	   LDFLAGS += -Wl,--no-as-needed
 	endif
-	
+
 	gccversion = $(shell $(CXX) --version)
 	gccisclang = $(findstring clang,$(gccversion))
 endif
@@ -354,8 +339,8 @@ endif
 endif
 
 ifeq ($(KERNEL),Darwin)
-	CXXFLAGS += -arch $(arch) -mmacosx-version-min=10.13
-	LDFLAGS += -arch $(arch) -mmacosx-version-min=10.13
+	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
@@ -446,14 +431,14 @@ endif
 ifeq ($(avx512),yes)
 	CXXFLAGS += -DUSE_AVX512
 	ifeq ($(comp),$(filter $(comp),gcc clang mingw))
-		CXXFLAGS += -mavx512bw
+		CXXFLAGS += -mavx512f -mavx512bw
 	endif
 endif
 
-ifeq ($(sse42),yes)
-	CXXFLAGS += -DUSE_SSE42
+ifeq ($(vnni),yes)
+	CXXFLAGS += -DUSE_VNNI
 	ifeq ($(comp),$(filter $(comp),gcc clang mingw))
-		CXXFLAGS += -msse4.2
+		CXXFLAGS += -mavx512vnni -mavx512dq -mavx512vl
 	endif
 endif
 
@@ -471,10 +456,10 @@ ifeq ($(ssse3),yes)
 	endif
 endif
 
-ifeq ($(sse3),yes)
-	CXXFLAGS += -DUSE_SSE3
+ifeq ($(mmx),yes)
+	CXXFLAGS += -DUSE_MMX
 	ifeq ($(comp),$(filter $(comp),gcc clang mingw))
-		CXXFLAGS += -msse3
+		CXXFLAGS += -mmmx
 	endif
 endif
 
@@ -509,6 +494,11 @@ 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
 		LDFLAGS += $(CXXFLAGS)
@@ -554,17 +544,16 @@ help:
 	@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-sse42            > x86 64-bit with sse42 support"
-	@echo "x86-64-modern           > x86 64-bit with sse41 support (x86-64-sse41)"
-	@echo "x86-64-sse41            > x86 64-bit with sse41 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-ssse3            > x86 64-bit with ssse3 support"
 	@echo "x86-64-sse3-popcnt      > x86 64-bit with sse3 and popcnt support"
-	@echo "x86-64-sse3             > x86 64-bit with sse3 support"
 	@echo "x86-64                  > x86 64-bit generic"
-	@echo "x86-32                  > x86 32-bit (also enables SSE)"
+	@echo "x86-32                  > x86 32-bit (also enables MMX and SSE)"
 	@echo "x86-32-old              > x86 32-bit fall back for old hardware"
 	@echo "ppc-64                  > PPC 64-bit"
 	@echo "ppc-32                  > PPC 32-bit"
@@ -603,7 +592,7 @@ help:
 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)
@@ -644,7 +633,7 @@ 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
 
 default:
@@ -669,13 +658,12 @@ config-sanity:
 	@echo "prefetch: '$(prefetch)'"
 	@echo "popcnt: '$(popcnt)'"
 	@echo "sse: '$(sse)'"
-	@echo "sse3: '$(sse3)'"
 	@echo "ssse3: '$(ssse3)'"
 	@echo "sse41: '$(sse41)'"
-	@echo "sse42: '$(sse42)'"
 	@echo "avx2: '$(avx2)'"
 	@echo "pext: '$(pext)'"
 	@echo "avx512: '$(avx512)'"
+	@echo "vnni: '$(vnni)'"
 	@echo "neon: '$(neon)'"
 	@echo ""
 	@echo "Flags:"
@@ -695,13 +683,12 @@ config-sanity:
 	@test "$(prefetch)" = "yes" || test "$(prefetch)" = "no"
 	@test "$(popcnt)" = "yes" || test "$(popcnt)" = "no"
 	@test "$(sse)" = "yes" || test "$(sse)" = "no"
-	@test "$(sse3)" = "yes" || test "$(sse3)" = "no"
 	@test "$(ssse3)" = "yes" || test "$(ssse3)" = "no"
 	@test "$(sse41)" = "yes" || test "$(sse41)" = "no"
-	@test "$(sse42)" = "yes" || test "$(sse42)" = "no"
 	@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"