merge support for 64-bit arm (the preprocessor tests had to be fixed) + benchmark on ipad air 2

merge from https://bitbucket.org/jpommier/pffft/src/default/

Signed-off-by: hayati ayguen <[email protected]>
diff --git a/pffft.c b/pffft.c
index 351d775..36bf6a8 100644
--- a/pffft.c
+++ b/pffft.c
@@ -57,6 +57,8 @@
   - 2011/10/02, version 1: This is the very first release of this file.
 */
 
+#include "pffft.h"
+
 /* detect compiler flavour */
 #if defined(_MSC_VER)
 #  define COMPILER_MSVC
@@ -68,7 +70,6 @@
 #  define _USE_MATH_DEFINES
 #endif
 
-#include "pffft.h"
 #include <stdlib.h>
 #include <stdio.h>
 #include <math.h>
@@ -156,7 +157,7 @@
 /*
   ARM NEON support macros
 */
-#elif !defined(PFFFT_SIMD_DISABLE) && defined(__arm__) 
+#elif !defined(PFFFT_SIMD_DISABLE) && (defined(__arm__) || defined(__aarch64__) || defined(__arm64__))
 #  include <arm_neon.h>
 typedef float32x4_t v4sf;
 #  define SIMD_SZ 4