[RUNTIME] Add flags for detecting user-defined Autotuner hooks (#5092)

When tracing user-defined Triton kernels in PyTorch 2, TorchDynamo
intercepts already instantiated `Autotuner` or `JITFunction` objects
([here](https://github.com/pytorch/pytorch/blob/ab42967238483b457af7228b9185b5816b83ed08/torch/_higher_order_ops/triton_kernel_wrap.py#L1034)).
This is because the decorator calls are normally made outside the
compiled part of the model code, with the latter covering only kernel
calls.

At this point, we don't support user-defined `Autotuner` `pre_hook` and
`post_hook` in PT2 compilation. However, due to the fact that the
`pre_hook` and `post_hook` attributes are always set to a lambda
([here](https://github.com/triton-lang/triton/blob/main/python/triton/runtime/autotuner.py#L55-L56)),
we can't detect that the hooks are set by the user by introspecting the
`Autotuner` object. This leads to silent incorrectness, as the
user-defined hooks are simply ignored.

This PR adds explicit boolean flags indicating whether the `pre_hook`
and / or `post_hook` are set by the user. With those in place, we can
detect this and raise a clear error around
[here](https://github.com/pytorch/pytorch/blob/ab42967238483b457af7228b9185b5816b83ed08/torch/_higher_order_ops/triton_kernel_wrap.py#L1072-L1074)
in TorchDynamo, avoiding silent incorrectness.

- [x] I am not making a trivial change, such as fixing a typo in a
comment.

- [x] I have written a PR description following these
[rules](https://cbea.ms/git-commit/#why-not-how).

- [x] I have run `pre-commit run --from-ref origin/main --to-ref HEAD`.

- Select one of the following.
  - [ ] I have added tests.
    - `/test` for `lit` tests
    - `/unittest` for C++ tests
    - `/python/test` for end-to-end tests
  - [x] This PR does not need a test because `just adding flags`.

- Select one of the following.
  - [x] I have not added any `lit` tests.
- [ ] The `lit` tests I have added follow these [best
practices](https://mlir.llvm.org/getting_started/TestingGuide/#filecheck-best-practices),
including the "tests should be minimal" section. (Usually running Python
code
    and using the instructions it generates is not minimal.)

Author: aakhundov

python/triton/runtime/autotuner.py

@@ -54,8 +54,11 @@ def __init__(
         # Hook to reset or restore for required tensors
         self.pre_hook = lambda kwargs, reset_only=False: 0
         self.post_hook = lambda kwargs, exception: 0
+        self.user_defined_pre_hook = False
+        self.user_defined_post_hook = False
         if pre_hook:
             self.pre_hook = pre_hook
+            self.user_defined_pre_hook = True
         elif (len(self.reset_to_zero) > 0 or len(self.restore_value) > 0):
 
             def _pre_hook(kwargs, reset_only=False):
@@ -68,6 +71,7 @@ def _pre_hook(kwargs, reset_only=False):
 
         if post_hook:
             self.post_hook = post_hook
+            self.user_defined_post_hook = True
         elif len(self.restore_value) > 0:
 
             def _post_hook(kwargs, exception):