Modified bli_kernel.h include order in blis.h.

Details:
- Delayed #include of bli_kernel.h in blis.h to prevent a situation where
  _kernel.h includes an optimized microkernel header, which uses BLIS types
  such as dim_t and inc_t, which would precede the definition of those types
  in bli_type_defs.h.
- Moved the #include of bli_kernel_macro_defs.h in bli_macro_defs.h to blis.h
  (immediately after that of bli_kernel.h).

frame/include/bli_macro_defs.h

@@ -66,7 +66,6 @@
 #include "bli_gentfunc_macro_defs.h"
 #include "bli_gentprot_macro_defs.h"
 
-#include "bli_kernel_macro_defs.h"
 #include "bli_mem_macro_defs.h"
 #include "bli_pool_macro_defs.h"
 #include "bli_obj_macro_defs.h"

frame/include/blis.h

@@ -64,11 +64,6 @@ extern "C" {
 #include <time.h>
 
 
-// -- BLIS kernel definitions --
-
-#include "bli_kernel.h"
-
-
 // -- BLIS definitions --
 
 #include "bli_type_defs.h"
@@ -76,6 +71,12 @@ extern "C" {
 #include "bli_extern_defs.h"
 
 
+// -- BLIS kernel definitions --
+
+#include "bli_kernel.h"
+#include "bli_kernel_macro_defs.h"
+
+
 // -- Base operation prototypes --
 
 #include "bli_init.h"