always use USE_SSE3_LDDQU on Windows+MSVC, and also on GCC when __SSE3__ is defined

src/vectormath/sse/quat_aos.h

@@ -530,7 +530,7 @@ VECTORMATH_FORCE_INLINE const Quat select( const Quat &quat0, const Quat &quat1,
 
 VECTORMATH_FORCE_INLINE void loadXYZW(Quat& quat, const float* fptr)
 {
-#ifdef USE_SSE2_LDDQU
+#ifdef USE_SSE3_LDDQU
 	quat = Quat(	SSEFloat(_mm_lddqu_si128((const __m128i*)((float*)(fptr)))).m128		);
 #else
 	SSEFloat fl;

src/vectormath/sse/vec_aos.h

@@ -331,7 +331,7 @@ VECTORMATH_FORCE_INLINE __m128 Vector3::get128( ) const
 
 VECTORMATH_FORCE_INLINE void loadXYZ(Vector3& vec, const float* fptr)
 {
-#ifdef USE_SSE2_LDDQU
+#ifdef USE_SSE3_LDDQU
 	vec = Vector3(	SSEFloat(_mm_lddqu_si128((const __m128i*)((float*)(fptr)))).m128 );
 #else
 	SSEFloat fl;
@@ -340,7 +340,7 @@ VECTORMATH_FORCE_INLINE void loadXYZ(Vector3& vec, const float* fptr)
 	fl.f[2] = fptr[2];
 	fl.f[3] = fptr[3];
     vec = Vector3(	fl.m128);
-#endif //USE_SSE2_LDDQU
+#endif //USE_SSE3_LDDQU
 	
 }
 

src/vectormath/sse/vectormath_aos.h

@@ -41,6 +41,8 @@
 #define Matrix3Ref Matrix3&
 
 #if (defined (_WIN32) && (_MSC_VER) && _MSC_VER >= 1400)
+	#define USE_SSE3_LDDQU
+
 	#define VM_ATTRIBUTE_ALIGNED_CLASS16(a) __declspec(align(16)) a
 	#define VM_ATTRIBUTE_ALIGN16 __declspec(align(16))
 	#define VECTORMATH_FORCE_INLINE __forceinline 
@@ -48,14 +50,15 @@
 	#define VM_ATTRIBUTE_ALIGNED_CLASS16(a) a __attribute__ ((aligned (16)))	
 	#define VM_ATTRIBUTE_ALIGN16 __attribute__ ((aligned (16)))	
 	#define VECTORMATH_FORCE_INLINE inline 
+	#ifdef __SSE3__
+		#define USE_SSE3_LDDQU
+	#endif //__SSE3__
 #endif//_WIN32
 
-#ifdef __SSE3__
-#define USE_SSE2_LDDQU
-#ifdef USE_SSE2_LDDQU
+
+#ifdef USE_SSE3_LDDQU
 #include <pmmintrin.h>//_mm_lddqu_si128
-#endif //USE_SSE2_LDDQU
-#endif //__SSE3__
+#endif //USE_SSE3_LDDQU
 
 
 // TODO: Tidy