[Mlir-commits] [mlir] [mlir][scf] Clamp normalized loop sizes where they are combined (PR #223256)

[llvmlistbot at llvm.org]

================
@@ -363,7 +374,8 @@ module attributes {transform.with_named_sequence} {
 //  CHECK-SAME:     , %[[ARG1:.+]]: index)
 //   CHECK-DAG:   %[[C0:.+]] = arith.constant 0 : index
 //   CHECK-DAG:   %[[C1:.+]] = arith.constant 1 : index
-//       CHECK:   scf.for %[[IV:.+]] = %[[C0]] to %[[ARG1]] step %[[C1]]
+//       CHECK:   %[[CL:.+]] = arith.maxsi %[[ARG1]], %[[C0]]
----------------
MaheshRavishankar wrote:

Can you clarify this a bit. If lb == -1, and ub == -2, whats the step. In this case if step is 1, then it should execute infinite iterations. if step == -1, then you have size = 1, and the loop runs once.

https://github.com/llvm/llvm-project/pull/223256