Avoid -Wduplicated-cond warnings in gdb/python

I tried building gdb with -Wduplicated-cond.  This patch fixes the
simpler issue that was found.

In Python 3, "int" and "long" are synonyms, so code like:

      else if (PyLong_Check (obj))
...
      else if (PyInt_Check (obj))

.... will trigger this warning.  The fix is to conditionalize the
PyInt_Check branches on Python 2.

Tested by rebuilding, with both version of Python, on x86-64 Fedora 24.

2016-09-20  Tom Tromey  <tom@tromey.com>

	* python/py-value.c (convert_value_from_python): Make PyInt_Check
	conditional on Python 2.
	* python/py-arch.c (archpy_disassemble): Make PyInt_Check
	conditional on Python 2.

gdb/ChangeLog

@@ -1,3 +1,10 @@
+2016-09-20  Tom Tromey  <tom@tromey.com>
+
+	* python/py-value.c (convert_value_from_python): Make PyInt_Check
+	conditional on Python 2.
+	* python/py-arch.c (archpy_disassemble): Make PyInt_Check
+	conditional on Python 2.
+
 2016-09-20  Edjunior Barbosa Machado  <emachado@linux.vnet.ibm.com>
 
 	* rs6000-tdep.c (ppc_process_record_op31): Fix record of Store String

gdb/python/py-arch.c

@@ -141,10 +141,12 @@ archpy_disassemble (PyObject *self, PyObject *args, PyObject *kw)
 	 conversion process.  */
       if (PyLong_Check (end_obj))
         end = PyLong_AsUnsignedLongLong (end_obj);
+#if PY_MAJOR_VERSION == 2
       else if (PyInt_Check (end_obj))
         /* If the end_pc value is specified without a trailing 'L', end_obj will
            be an integer and not a long integer.  */
         end = PyInt_AsLong (end_obj);
+#endif
       else
         {
           Py_DECREF (end_obj);

gdb/python/py-value.c

@@ -1642,6 +1642,7 @@ convert_value_from_python (PyObject *obj)
 	  else
 	    value = value_from_longest (builtin_type_pylong, l);
 	}
+#if PY_MAJOR_VERSION == 2
       else if (PyInt_Check (obj))
 	{
 	  long l = PyInt_AsLong (obj);
@@ -1649,6 +1650,7 @@ convert_value_from_python (PyObject *obj)
 	  if (! PyErr_Occurred ())
 	    value = value_from_longest (builtin_type_pyint, l);
 	}
+#endif
       else if (PyFloat_Check (obj))
 	{
 	  double d = PyFloat_AsDouble (obj);