[Mlir-commits] [mlir] [MLIR][Python] Improve Iterator performance. Don't `throw` in `dunderNext` methods. (PR #175377)

[llvmlistbot at llvm.org]

================
@@ -327,13 +331,15 @@ void PyBlockList::bind(nb::module_ &m) {
 nb::typed<nb::object, PyOpView> PyOperationIterator::dunderNext() {
   parentOperation->checkValid();
   if (mlirOperationIsNull(next)) {
-    throw nb::stop_iteration();
+    PyErr_SetNone(PyExc_StopIteration);
+    // python functions should return NULL after setting any exception
+    return nb::object();
   }
 
   PyOperationRef returnOperation =
       PyOperation::forOperation(parentOperation->getContext(), next);
   next = mlirOperationGetNextInBlock(next);
-  return returnOperation->createOpView();
+  return nb::cast(returnOperation->createOpView());
----------------
MaPePeR wrote:

Just noticed I dont need this `nb::cast`. Function already returned nb::typed<nb::object, PyOpView>` and `createOpView` returns `nb::object` already.

https://github.com/llvm/llvm-project/pull/175377