Summary:
# Problem
`TORCH_WARN` can cause massive log spam.
I output the logs for before and after adding this change.
*Before:*
* The log file size was ~61.15 MB(61148028 bytes).
*After:*
* The log filesize was ~56.44 MB(56444057) bytes.
# Context
Looks like we tried to land this change earlier but it was reverted:
* D59413413
* Reverted https://github.com/pytorch/pytorch/pull/130047 on behalf of https://github.com/clee2000 due to broke test_overrides.py::TestTorchFunctionWarning::test_warn_on_invalid_torch_function
# Testing Update
`test_warn_on_invalid_torch_function` would fail because the warning would not be called on the handling of the second torch function class since `TORCH_WARN_ONCE` stops repeats globally.
Updated so that it runs separate programs. (Was not able to actually run the test, could someone help me with that
Test Plan: Need help with this...
Differential Revision: D60561181
Pull Request resolved: https://github.com/pytorch/pytorch/pull/132374
Approved by: https://github.com/ezyang
https://github.com/pytorch/pytorch/issues/105290
The problem in the original flow is that:
(1) the user calls `torch.mul(complex_tensor, complex_scalar)
(2) python arg parser wraps the complex scalar in a `scalar_tensor`, and dispatches to `aten.mul.Tensor(self, scalar_other)`
(3) autograd sees `aten.mul.Tensor`, calls `scalar_other.conj()` [here](https://github.com/pytorch/pytorch/blob/main/torch/csrc/autograd/FunctionsManual.cpp#L597)
(4) during proxy tensor tracing, this gets dispatched to `aten._conj(scalar_tensor)`
(5) when we hit __torch_dispatch__, the scalar_tensor is converted back into a plain python scalar
(6) we error during tracing, because in `FunctionalTensorMode.__torch_dispatch__` we try to redispatch on `aten._conj.default(plain_python_scalar)`, and this overload does not accept python scalars.
My attempted fix in this PR is to update `TensorBase::conj()` to check if the current tensor is a scalar tensor (wrapped number), and if so, manually:
(1) convert the scalar tensor back into a scalar
(2) call scalar.conj() directly
(3) convert the result back into a wrapped tensor
This avoids having to go through python entirely in the tracing case (which is fine, because these scalar tensors are constants that we can const-prop during tracing anyway).
Notable, I did **not** add e.g. a new `aten._conj.Scalar` overload. This would not actually fix the problem, since the bug is that we call `aten._conj.default(python_scalar)` directly. we would also need to muck with all `__torch_dispatch__` call sites to know to convert python scalars back into tensors directly.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/131482
Approved by: https://github.com/zou3519, https://github.com/ezyang
ghstack dependencies: #131403
Fix static `py::object`s with `py::gil_safe_call_once_and_store`.
The following code will leak a `py::object` which will call its destructor when shutdown the program. The destructor will call `Py_DECREF(obj.m_ptr)` which may raise a segmentation fault.
```c++
void func() {
static py::object obj = py::module_::import("foo").attr("bar");
...
}
```
The correct code is to use raw pointers rather than the instance.
```c++
void func() {
static py::object* obj_ptr = new py::object{py::module_::import("foo").attr("bar")};
py::object obj = *obj_ptr;
...
}
```
This PR uses the `py::gil_safe_call_once_and_store` function from `pybind11`, which can run arbitrary initialization code only once under the Python GIL thread safely.
```c++
void func() {
PYBIND11_CONSTINIT static py::gil_safe_call_once_and_store<py::object> storage;
py::object obj = storage
.call_once_and_store_result(
[]() -> py::object {
return py::module_::import("foo").attr("bar");
}
)
.get_stored();
...
}
```
Pull Request resolved: https://github.com/pytorch/pytorch/pull/130341
Approved by: https://github.com/ezyang, https://github.com/malfet
We add torch.library.Library._register_torch_dispatch_rule. Here, a user
can provide us a specific rule to run for a specific
(torch_dispatch_class, operator) pair. The motivation is that a user
might want to extend a subclass/mode but may not have access to the
source code of the subclass/mode.
I'll make this public in a follow-up PR if we think the approach and API
is good.
Keep in mind that many subclasses will likely deliver their own open
registration solution (DTensor has register_sharding_prop_rule and NJT
has register_jagged_op); _register_torch_dispatch_rule is meant as a
catch-all open registration mechanism for when the subclass hasn't
provided anything more specific.
Test Plan:
- new tests
Pull Request resolved: https://github.com/pytorch/pytorch/pull/130064
Approved by: https://github.com/albanD
We add torch.library.Library._register_torch_dispatch_rule. Here, a user
can provide us a specific rule to run for a specific
(torch_dispatch_class, operator) pair. The motivation is that a user
might want to extend a subclass/mode but may not have access to the
source code of the subclass/mode.
I'll make this public in a follow-up PR if we think the approach and API
is good.
Keep in mind that many subclasses will likely deliver their own open
registration solution (DTensor has register_sharding_prop_rule and NJT
has register_jagged_op); _register_torch_dispatch_rule is meant as a
catch-all open registration mechanism for when the subclass hasn't
provided anything more specific.
Test Plan:
- new tests
Pull Request resolved: https://github.com/pytorch/pytorch/pull/130064
Approved by: https://github.com/albanD
We add torch.library.Library._register_torch_dispatch_rule. Here, a user
can provide us a specific rule to run for a specific
(torch_dispatch_class, operator) pair. The motivation is that a user
might want to extend a subclass/mode but may not have access to the
source code of the subclass/mode.
I'll make this public in a follow-up PR if we think the approach and API
is good.
Keep in mind that many subclasses will likely deliver their own open
registration solution (DTensor has register_sharding_prop_rule and NJT
has register_jagged_op); _register_torch_dispatch_rule is meant as a
catch-all open registration mechanism for when the subclass hasn't
provided anything more specific.
Test Plan:
- new tests
Pull Request resolved: https://github.com/pytorch/pytorch/pull/130064
Approved by: https://github.com/albanD
This PR enables the misc-XX checks in clang-tidy. Meanwhile, I excluded some of them that require a lot of code changes and have no immediate benefits. Some additional fixes and suppression were also given.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/110283
Approved by: https://github.com/albanD
**Update:** Made refactor of the original PR. See the original description below, but here I'll describe the updates:
(1) TLS changes in `TorchDispatchModeTLS.h/cpp`.
I added a `TorchDispatchModeKey` enum, that (for now) just contains PROXY and FAKE. The ModeTLS used to just contain a `std::vector<std::shared_ptr<c10::SafePyObject>>` corresponding to the mode stack. It now **also** contains a separate array of "infra modes", indexed by mode key (PROXY and FAKE, with a new addition, FUNCTIONAL, coming later in the stack).
`TorchDispatchModeTLS::push_onto_stack` and `TorchDispatchModeTLS::pop_stack` are now a bit more complicated. Pushing accepts an optional mode_key, which if set, tells us to add the given mode directly to our "infra_modes" array. Popping will first check the "user mode" stack, before trying to pop anything from the infra mode stack. It also optionally returns the mode key of the mode we popped if there was one - that way if we push that same mode back onto the TLS later, we know where it goes.
`TorchDispatchModeTLS::dispatch_mode_enabled()` now accepts an optional `skip_infra_modes` param, so you can separately query if there are "any modes at all", or if there are "any user modes".
`TorchDispatchModeTLS::get/set/unset_mode()` all take in a mode key, and get/set/unset the mode at that particular mode key (meaning they are only meant to be used for infra modes).
There were also some mild codegen changes to support the new enum
(2) `fake_tensor.py/proxy_tensor.py/_python_dispatch.py`
The way I tell the infra that certain subclasses/modes are "infra" is through the enum: I gave `FakeTensor` and `FakeTensorMode` a `self._mode_key = torch._C.TorchDispatchModeKey.FAKE`. `TorchDispatchMode.__enter/exit__()` (in `_python_dispatch.py` now check if the current mode has a mode key, and if so they plumb it into any `push_onto_stack()` calls (which eventually instructs `TorchDispatchModeTLS` where to put the mode). Same thing for `ProxyTorchDispatchMode`.
I also had to change both of these mode's enter/exit, to handle the fact that there can no longer be multiple proxy/fake modes on the mode stack at once. I updated them both to have a `self.enter_stack: List[Optional[TorchDispatchMode]]` - whenever we push a given mode in `__enter__`, we remove the current ambient fake/proxy mode from the mode stack, and save it in `enter_stack`, so that on exit we can reset the state properly.
(2) dispatching logic in `python_arg_parser.cpp`
This is where the core dispatching logic changes are. I added two helpers, `dispatch_on_subclass()` and `dispatch_on_mode()`. The overall dispatching order is now:
```
(a) dispatch_on_mode() # try user modes first (where the mode stack automatically considers infra modes last)
(b) dispatch_on_subclass() # try user subclasses next (skipping infra subclasses)
(c) dispatch_on_subclass() # try infra subclasses next (skipping user subclasses)
```
Note that we still want "user subclasses" to run before "infra modes". As Ed helped me realize, this will work today: If proxy/fake modes in step 1, they'll return NotImplemented if they see a user subclass, allowing us to redispatch to the user subclass.
How do (b) and (c) distinguish between user and infra subclasses? Infra subclasses (FakeTensor, and later FunctionalTensor) are required to have a `_mode_key` hidden on the subclass - so we filter via arguments that do/don't have the _mode_key.
(3) I also changed `DoubleTensor` to `TwoTensor` to minimize confusion (@albanD pointed out that DoubleTensor would be easily confused with `torch.FloatTensor` and friends).
----- original description below -----
The main purpose of this PR is to fix the "ordering problem" between torch_dispatch modes, where we want to ensure that our Fake and Proxy dispatch modes always run **after** any dispatch modes created by the user, regardless of where they are in the stack. See this doc for more details: https://docs.google.com/document/d/1COQ291nOZvtFnzGTQMJqoYZ3sttEYFw_7HbfSyL8gcA/edit
Full set of changes below. I ended up including a few semi-related changes in this PR that I documented - but if folks would rather I separate them out, happy to try to do that.
**(1) Add dedicated TLS slots for FakeTensorMode and ProxyTensorMode**
This is the main component of this PR. There are two new slots, `TorchDispatchModeTLS.fake_mode_` and `TorchDispatchModeTLS.proxy_mode_`, which correspond to a single "global" fake and proxy mode. There is now an invariant that `torchDispatchModeState.stack_` can never contain either of these modes.
I also added a `TorchDispatchModeTLS::maybe_highest_mode()` helper that consults the `stack_` as well as both the proxy and fake slots, and returns the highest priority mode - this is because there are a few places in the codebase where we legitimately want to get the highest priority mode, *including* fake or proxy, if one is set.
This also made the implementations of the existing `disable_proxy_modes_tracing()` and `get_innermost_proxy_mode()` marginally simpler.
**(2) Updated the dispatching logic in handle_torch_function_no_python_arg_parser()**
This is the function that actually figures out which torch_dispatch implementation to call, given the current mode stack and tensor subclass inputs. This function got marginally more complicated as part of the refactor: First we inspect the mode stack and any non-fake subclass inputs. Then we check for the proxy mode slot. Then we check for the Fake mode slot, before finally checking for any fake subclass inputs.
**(3) new python `_get_fake_tensor_mode()` and `_get_proxy_tensor_mode()` API's**
Before, if you wanted to see if proxy or fake modes were active in python, you would have to consult the mode stack. Since these two modes are no longer part of the actual mode stack, I added two new API's to directly check if either proxy or fake modes are active.
**(4) Allow traceable tensor subclasses to access storages from python**
This is convenient later in the stack, where AOTAutograd needs to detect aliasing of inputs and outputs, where those inputs and outputs might be tensor subclasses. Previously, `x.untyped_storage()` would raise an error if `x` was a subclass. In this PR, I tried to relax this constraint as little as possible: `THPVariable_storage()` will only try to return a storage to python if the tensor subclass that you are passing in is "traceable"
**(5) Fixed subclass fakeification**
@wanchaol recently added support to be able to fakeify tensor subclasses. That fakeification logic works in most cases, but there is one case it doesn't handle: autograd metadata. In particular, since autograd sees our tensor subclasses and not their desugared tensors, we need to make sure that our fakeified subclass has the same autograd metadata as the original subclass. I updated `meta_utils.py` to make sure that the autograd metadata is correct.
**(6) make tensor subclasses resizeable**
Previously we didn't allow tensor subclasses to be resizeable. I ran into an issue where fakeifying a tensor subclass occasionally requires swapping out its storage, which can involve resizing the tensor. Mechanically, this required updating `at::for_blob()` to expose a way to request that the tensor that you create has resizeable storage, and then using this new API in `_make_wrapper_tensor()`.
**(7) Added a basic DoubleTensor subclass for testing**
I use this subclass more later in this stack in my AOTAutograd tests - but it serves as a simple subclass example to test the dispatch ordering in this PR.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/104482
Approved by: https://github.com/ezyang
ghstack dependencies: #107415
This moves the `overloaded_args` field from FunctionSignature to PythonArgs. FunctionSignature is shared by all calls and should be immutable. PythonArgs contains the parsing results for an single call to the PyTorch API.
I did not measure a difference in performance in the "overrides_benchmark", although I expect there to be a bit more work in the common case. Note that the noise factor for the benchmark is much larger than the differences reported below:
Before:
```
Type tensor had a minimum time of 2.3615360260009766 us and a standard deviation of 0.7833134150132537 us.
Type SubTensor had a minimum time of 10.473251342773438 us and a standard deviation of 0.1973132457351312 us.
Type WithTorchFunction had a minimum time of 5.484819412231445 us and a standard deviation of 0.13305981701705605 us.
Type SubWithTorchFunction had a minimum time of 11.098146438598633 us and a standard deviation of 0.15598918253090233 us.
```
After:
```
Type tensor had a minimum time of 2.2134780883789062 us and a standard deviation of 0.802064489107579 us.
Type SubTensor had a minimum time of 10.625839233398438 us and a standard deviation of 0.15155907021835446 us.
Type WithTorchFunction had a minimum time of 5.520820617675781 us and a standard deviation of 0.23115111980587244 us.
Type SubWithTorchFunction had a minimum time of 11.227846145629883 us and a standard deviation of 0.23032321769278497 us.
```
Fixes#106974
Pull Request resolved: https://github.com/pytorch/pytorch/pull/106983
Approved by: https://github.com/zou3519, https://github.com/ezyang, https://github.com/albanD
before the PR, when compiling a function with signature symint/symintlist/intlist, we have runtime error like ```argument 'shifts' must be tuple of ints, not FakeTensor```. see newly added unit test in test/dynamo/test_misc.py for repro
after the PR, for FakeTensor with empty size and numel()=1, we will try
to convert it into symint/symintlist. we will likely see expected
exception
```torch._subclasses.fake_tensor.DataDependentOutputException / aten._local_scalar_dense.default``` during conversion
reference PR:
* we handle FakeTensor for symintlist as 1st varags: https://github.com/pytorch/pytorch/pull/97508
* we handle FakeTensor for intlist in a similar way:
https://github.com/pytorch/pytorch/pull/85759/files
* call local_scalar_dense on a FakeTensor:
f7365eca90
Pull Request resolved: https://github.com/pytorch/pytorch/pull/103448
Approved by: https://github.com/yanboliang
Failing mechanism on #95424 :
In dynamo mode, when passing numpy.int_ to 'shape' like param (Sequence[Union[int, symint]]) is wrapped as list with FakeTensor. However, in python_arg_parser, parser expect int in symint_list but got FakeTensor.
Following #85759, this PR allow tensor element in symint_list when in dynamo mode
This PR also fix below test with similar failing mechanism
pytest ./generated/test_huggingface_diffusers.py -k test_016
pytest ./generated/test_ustcml_RecStudio.py -k test_036
Fixes #ISSUE_NUMBER
Pull Request resolved: https://github.com/pytorch/pytorch/pull/97508
Approved by: https://github.com/yanboliang
When investigating failures in https://github.com/pytorch/pytorch/pull/100017 I realized that we were reentering FakeTensorMode even though there was already one on the stack. Although we have attempted assert for these cases in the past, e.g., as in https://github.com/pytorch/pytorch/pull/97186 it seems that the existing protections were insufficient.
In this particular case, the reapplication of FakeTensorMode was due to an interaction with NotImplemented multiple dispatch handling. If proxy tensor mode detects an unrecognized tensor type (this includes FakeTensor, if it is not tracked with a proxy), it will return NotImplemented to give this tensor a chance to unpack itself into proxyable operation. However, this is never the right thing for FakeTensor, where no unpacking is possible. However, today, FakeTensor attempts to reapply the FakeTensorMode, resulting in FakeTensorMode being twice on the stack.
This PR does a number of things:
* It adds an assert in `FakeTensorMode.__torch_dispatch__` that you must not already have this mode on the stack, this is ALWAYS an error
* It modifies `FakeTensor.__torch_dispatch__` to return `NotImplemented` if the mode is already active. This prevents us from readding the mode on the stack
* It adds a new logging artifact `not_implemented` which you can use to get debug logs about all of the times a `__torch_dispatch__` handler returned NotImplemented and why it did so. Your subclass has to manually opt into this logging, but I inserted the necessary logs for ProxyTensorMode and FakeTensor(Mode)
* `with fake_mode` now no-ops if the fake mode is already on the stack, which is what users want anyway
* I am BREAKING pre-autograd tracing, because it is currently doing something weird with the original C++ mode stack. Brian is going to follow up with a fix next week.
Signed-off-by: Edward Z. Yang <ezyang@meta.com>
Pull Request resolved: https://github.com/pytorch/pytorch/pull/102091
Approved by: https://github.com/thiagocrepaldi, https://github.com/eellison, https://github.com/wanchaol, https://github.com/bdhirsh
This PR introduces **-Wmissing-prototypes** of clang-tidy to prevent further coding errors such as the one fixed by PR #96714.
<!--
copilot:summary
-->
### <samp>🤖 Generated by Copilot at fd2cf2a</samp>
This pull request makes several internal functions static to improve performance and avoid name clashes. It also fixes some typos, formatting, and missing includes in various files. It adds a new .clang-tidy check to warn about missing prototypes for non-static functions.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/96805
Approved by: https://github.com/malfet, https://github.com/albanD
