n^-m gives float result (instead of error)

lcode.c

@@ -1,5 +1,5 @@
 /*
-** $Id: lcode.c,v 2.86 2014/04/01 14:39:55 roberto Exp roberto $
+** $Id: lcode.c,v 2.87 2014/04/02 16:44:42 roberto Exp roberto $
 ** Code generator for Lua
 ** See Copyright Notice in lua.h
 */
@@ -768,8 +768,6 @@ static int validop (int op, TValue *v1, TValue *v2) {
       return (tointeger(v1, &i) && tointeger(v2, &i));
     case LUA_OPMOD:  /* integer module by 0 */
       return !(ttisinteger(v1) && ttisinteger(v2) && ivalue(v2) == 0);
-    case LUA_OPPOW:  /* negative integer exponentiation */
-      return !(ttisinteger(v1) && ttisinteger(v2) && ivalue(v2) < 0);
     default: return 1;  /* everything else is valid */
   }
 }