From ba3c333611c4a825ae30ea97484b937e88b88778 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Przemys=C5=82aw=20Pawe=C5=82czyk?= <przemoc@gmail.com>
Date: Tue, 5 Sep 2023 13:58:47 +0200
Subject: [PATCH] make : improve cpuinfo handling on x86 hosts (#1238)

* make : simplify and correct x86 ISA extensions detection on the host

It got broken in commit c5f9acf4b797 for Haiku and Mac OS (Intel),
which report CPU features in upper case.

Now we're finding the names in case-insensitive manner and as words.
SSE3 detection has been corrected for Linux, which uses PNI for that
(Prescott New Instructions).

* make : use dmesg.boot in FreeBSD/DragonFlyBSD to detect x86 ISA extensions on the host

* make : enable x86 ISA extensions on the host both in CFLAGS and CXXFLAGS

* make : correct AVX x86 ISA extension detection on macOS (Intel) host

It got broken in commit c5f9acf4b797.  macOS calls it AVX1.0.
---
 Makefile | 60 ++++++++++++++++++++++++++++----------------------------
 1 file changed, 30 insertions(+), 30 deletions(-)

diff --git a/Makefile b/Makefile
index a2631011..d775b74b 100644
--- a/Makefile
+++ b/Makefile
@@ -65,57 +65,57 @@ endif
 # Architecture specific
 # TODO: probably these flags need to be tweaked on some architectures
 #       feel free to update the Makefile for your architecture and send a pull request or issue
-ifeq ($(UNAME_M),$(filter $(UNAME_M),x86_64 i686))
+ifeq ($(UNAME_M),$(filter $(UNAME_M),x86_64 i686 amd64))
 	ifeq ($(UNAME_S),Darwin)
 		CPUINFO_CMD := sysctl machdep.cpu.features
 	else ifeq ($(UNAME_S),Linux)
 		CPUINFO_CMD := cat /proc/cpuinfo
 	else ifneq (,$(filter MINGW32_NT% MINGW64_NT%,$(UNAME_S)))
 		CPUINFO_CMD := cat /proc/cpuinfo
+	else ifneq (,$(filter DragonFly FreeBSD,$(UNAME_S)))
+		CPUINFO_CMD := grep Features /var/run/dmesg.boot
 	else ifeq ($(UNAME_S),Haiku)
 		CPUINFO_CMD := sysinfo -cpu
 	endif
 
-	ifdef CPUINFO_CMD  	
-    AVX_M := $(shell $(CPUINFO_CMD) | grep -m 1 "avx ")
-		ifneq (,$(findstring avx,$(AVX_M)))
-			CFLAGS += -mavx
-		endif
-    
-		AVX2_M := $(shell $(CPUINFO_CMD) | grep -m 1 "avx2 ")
-		ifneq (,$(findstring avx2,$(AVX2_M)))
-			CFLAGS += -mavx2
+	ifdef CPUINFO_CMD
+		AVX_M := $(shell $(CPUINFO_CMD) | grep -iwE 'AVX|AVX1.0')
+		ifneq (,$(AVX_M))
+			CFLAGS   += -mavx
+			CXXFLAGS += -mavx
 		endif
 
-		FMA_M := $(shell $(CPUINFO_CMD) | grep -m 1 "fma ")
-		ifneq (,$(findstring fma,$(FMA_M)))
-			CFLAGS += -mfma
+		AVX2_M := $(shell $(CPUINFO_CMD) | grep -iw 'AVX2')
+		ifneq (,$(AVX2_M))
+			CFLAGS   += -mavx2
+			CXXFLAGS += -mavx2
 		endif
 
-		F16C_M := $(shell $(CPUINFO_CMD) | grep -m 1 "f16c ")
-		ifneq (,$(findstring f16c,$(F16C_M)))
-			CFLAGS += -mf16c
-
-			AVX1_M := $(shell $(CPUINFO_CMD) | grep -m 1 "avx ")
-			ifneq (,$(findstring avx,$(AVX1_M)))
-				CFLAGS += -mavx
-			endif
+		FMA_M := $(shell $(CPUINFO_CMD) | grep -iw 'FMA')
+		ifneq (,$(FMA_M))
+			CFLAGS   += -mfma
+			CXXFLAGS += -mfma
 		endif
 
-		SSE3_M := $(shell $(CPUINFO_CMD) | grep -m 1 "sse3 ")
-		ifneq (,$(findstring sse3,$(SSE3_M)))
-			CFLAGS += -msse3
+		F16C_M := $(shell $(CPUINFO_CMD) | grep -iw 'F16C')
+		ifneq (,$(F16C_M))
+			CFLAGS   += -mf16c
+			CXXFLAGS += -mf16c
 		endif
 
-		SSSE3_M := $(shell $(CPUINFO_CMD) | grep -m 1 "ssse3 ")
-		ifneq (,$(findstring ssse3,$(SSSE3_M)))
-			CFLAGS += -mssse3
+		SSE3_M := $(shell $(CPUINFO_CMD) | grep -iwE 'PNI|SSE3')
+		ifneq (,$(SSE3_M))
+			CFLAGS   += -msse3
+			CXXFLAGS += -msse3
+		endif
+
+		SSSE3_M := $(shell $(CPUINFO_CMD) | grep -iw 'SSSE3')
+		ifneq (,$(SSSE3_M))
+			CFLAGS   += -mssse3
+			CXXFLAGS += -mssse3
 		endif
 	endif
 endif
-ifeq ($(UNAME_M),amd64)
-	CFLAGS += -mavx -mavx2 -mfma -mf16c
-endif
 
 ifneq ($(filter ppc64%,$(UNAME_M)),)
 	POWER9_M := $(shell grep "POWER9" /proc/cpuinfo)