[RFC] use templated isl types isPromotableToRegistersBelow

Templated isl types require the user to specify the domain and range
universes of isl objects, allowing the compiler to check whether
it makes sense to combine pairs of objects.
This RFC only converts isPromotableToRegistersBelow and some related
functions to illustrate the effect.
The isPromotableToRegistersBelow was already applying operations
correctly, so the code itself did not require any changes.
However, one variable was reused to store different types
of intermediate result and this one had to be split up into
several variables because they now have different types.

Author: Sven Verdoolaege

tc/core/polyhedral/cuda/memory_promotion_heuristic.cc

@@ -370,23 +370,25 @@ bool accessSubscriptsAreUnrolledLoops(
  * thread associated to a given pair of tensor element and outer schedule
  * iteration.
  */
+template <typename Outer>
 bool isPromotableToRegistersBelow(
     const TensorReferenceGroup& group,
     const detail::ScheduleTree* root,
     const detail::ScheduleTree* scope,
-    isl::multi_union_pw_aff outer,
-    isl::multi_union_pw_aff thread) {
+    isl::MultiUnionPwAff<Statement, Outer> outer,
+    isl::MultiUnionPwAff<Statement, Thread> thread) {
   if (!accessSubscriptsAreUnrolledLoops(
-          group, root, scope, outer.flat_range_product(thread))) {
+          group, root, scope, outer.range_product(thread))) {
     return false;
   }
 
   auto originalAccesses = group.originalAccesses();
-  auto map = isl::union_map::from(outer);
-  map = map.range_product(originalAccesses);
-  map = map.apply_domain(isl::union_map::from(thread));
+  auto outerMap = isl::UnionMap<Statement, Outer>::from(outer);
+  auto pair = outerMap.range_product(originalAccesses);
+  auto threadMap = isl::UnionMap<Statement, Thread>::from(thread);
+  auto threadToPair = pair.apply_domain(threadMap);
 
-  return map.is_injective();
+  return threadToPair.is_injective();
 }
 
 /*