Use NEON on ARM hardware via sse2neon.h

The autoconf changes were adapted from:
https://github.com/glennrp/libpng/blob/libpng16/configure.ac

src/vector_math.h

@@ -8,8 +8,10 @@
 
 #define ks_lt_index(a, b) ((a).value < (b).value)
 
-#ifdef  USE_SSE
+#if   defined(INTEL_SSE)
 #include <emmintrin.h>
+#elif defined(ARM_NEON)
+#include "sse2neon.h"
 #endif
 
 /*
@@ -338,7 +340,7 @@
 
 
 
-#ifdef USE_SSE
+#if defined(INTEL_SSE) || defined(ARM_NEON)
 /*
 From https://github.com/herumi/fmath/blob/master/fastexp.cpp
 
@@ -524,4 +526,4 @@ static inline void remez9_0_log2_sse(double *values, size_t num)
 
 
 
-#endif
+#endif