Use ParamInfo instead of NamedTuple is annotation

This required moving ParamInfo definition from class scope
to module scope, since referencing Kernel.ParamInfo from
annotations of methods of the Kernel class results in error
that Kernel class does not yet exist.

Author: oleksandr-pavlyk

cuda_core/cuda/core/experimental/_module.py

@@ -3,7 +3,7 @@
 # SPDX-License-Identifier: Apache-2.0
 
 from collections import namedtuple
-from typing import NamedTuple, Optional, Union
+from typing import Optional, Union
 from warnings import warn
 
 from cuda.core.experimental._utils.clear_error_support import (
@@ -184,6 +184,9 @@ def cluster_scheduling_policy_preference(self, device_id: int = None) -> int:
         )
 
 
+ParamInfo = namedtuple("ParamInfo", ["offset", "size"])
+
+
 class Kernel:
     """Represent a compiled kernel that had been loaded onto the device.
 
@@ -196,7 +199,6 @@ class Kernel:
     """
 
     __slots__ = ("_handle", "_module", "_attributes")
-    ParamInfo = namedtuple("ParamInfo", ["offset", "size"])
 
     def __new__(self, *args, **kwargs):
         raise RuntimeError("Kernel objects cannot be instantiated directly. Please use ObjectCode APIs.")
@@ -218,7 +220,7 @@ def attributes(self) -> KernelAttributes:
             self._attributes = KernelAttributes._init(self._handle)
         return self._attributes
 
-    def _get_arguments_info(self, param_info=False) -> tuple[int, list[NamedTuple]]:
+    def _get_arguments_info(self, param_info=False) -> tuple[int, list[ParamInfo]]:
         attr_impl = self.attributes
         if attr_impl._backend_version != "new":
             raise NotImplementedError("New backend is required")
@@ -229,7 +231,7 @@ def _get_arguments_info(self, param_info=False) -> tuple[int, list[NamedTuple]]:
             if result[0] != driver.CUresult.CUDA_SUCCESS:
                 break
             if param_info:
-                p_info = Kernel.ParamInfo(offset=result[1], size=result[2])
+                p_info = ParamInfo(offset=result[1], size=result[2])
                 param_info_data.append(p_info)
             arg_pos = arg_pos + 1
         if result[0] != driver.CUresult.CUDA_ERROR_INVALID_VALUE:
@@ -243,8 +245,8 @@ def num_arguments(self) -> int:
         return num_args
 
     @property
-    def arguments_info(self) -> list[NamedTuple]:
-        """list[NamedTuple[int, int]]: (offset, size) for each argument of this function"""
+    def arguments_info(self) -> list[ParamInfo]:
+        """list[ParamInfo]: (offset, size) for each argument of this function"""
         _, param_info = self._get_arguments_info(param_info=True)
         return param_info