needforspeed: added Py_LOCAL macro, based on the LOCAL macro used

for SRE and others.  applied Py_LOCAL to relevant portion of ceval,
which gives a 1-2% speedup on my machine.  ymmv.

Include/pyport.h

@@ -137,6 +137,23 @@ typedef Py_intptr_t	Py_ssize_t;
 #   endif
 #endif
 
+/* PY_LOCAL can be used instead of static to get the fastest possible calling
+ * convention for functions that are local to a given module.  It also enables
+ * inlining, where suitable. */
+
+#undef USE_INLINE /* XXX - set via configure? */
+
+#if defined(_MSC_VER)
+ /* ignore warnings if the compiler decides not to inline a function */ 
+#pragma warning(disable: 4710)
+/* fastest possible local call under MSVC */
+#define Py_LOCAL(type) static __inline type __fastcall
+#elif defined(USE_INLINE)
+#define Py_LOCAL(type) static inline type
+#else
+#define Py_LOCAL(type) static type
+#endif
+
 #include <stdlib.h>
 
 #include <math.h> /* Moved here from the math section, before extern "C" */