[LINUX] Add compiler flags to fix floating point precision error (Har…

…bourMasters#399)

These flags are known to fix one known issue exclusive to Linux: the Volvagia boss battle sequence. The softlock that occurs points towards a possible floating point precision error, possibly tied to the camera/Volvagia movement. This does not occur for the Windows build. It's possible that there are more issues that gets fixed by these flags. These flags will ensure that the compiler follows the IEEE 754 standard, which so happens to be the same behaviour that Windows uses. For more details, read this informative stackoverflow post: https://stackoverflow.com/a/16395650

soh/Makefile

@@ -18,8 +18,8 @@ WARN := \
 	-funsigned-char \
 	-m32 -mhard-float -fno-stack-protector -fno-common -fno-zero-initialized-in-bss -fno-strict-aliasing -fno-inline-functions -fno-inline-small-functions -fno-toplevel-reorder -ffreestanding -fwrapv \
 
-CXXFLAGS := $(WARN) -std=c++20 -D_GNU_SOURCE -fpermissive -no-pie -nostdlib -march=i386
-CFLAGS := $(WARN) -std=c99 -D_GNU_SOURCE -no-pie -nostdlib -march=i386
+CXXFLAGS := $(WARN) -std=c++20 -D_GNU_SOURCE -fpermissive -no-pie -nostdlib -march=i386 -msse2 -mfpmath=sse
+CFLAGS := $(WARN) -std=c99 -D_GNU_SOURCE -no-pie -nostdlib -march=i386 -msse2 -mfpmath=sse
 LDFLAGS := -m32
 CPPFLAGS := -MMD