## `VariableTracker::build()` hides the Builders
### The problem
In the current code, creating a `VariableTracker` involves choosing one of two `Builder` classes and either calling a method, or calling a constructor that creates an object that you immediately call, [like this](083c9149b7/torch/_dynamo/variables/functions.py (L761-L768)).
Variations on this code are repeated in many places.
More, the `Builder` classes have a lot of dependencies, so they have to be loaded late in the whole import process to avoid circular imports, so they end up being repeatedly imported at local scope.
### The solution
In this commit, the import from `builder` and the logic of choosing and calling the Builder class are hidden in a single static factory method, `VariableTracker.build()`, easier to reason about and to import.
This commit net lowers the total lines of code by over 150 lines by removing repetitive logic and unnecessary local imports.
**CHANGES:** Originally the name of the static method was `VariableTracker.create()` but a static method on a derived class, `LazyVariableTracker.create()` now exists with a different signature that's irreconcilable, so the new static method was renamed to `VariableTracker.build()`.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/135714
Approved by: https://github.com/jansel
This adds Dynamo tracing support for the host-side Triton TMA API (see `create_2d_tma_descriptor` calls on the host in the [Triton tutorial](https://triton-lang.org/main/getting-started/tutorials/09-persistent-matmul.html#sphx-glr-getting-started-tutorials-09-persistent-matmul-py)). A few notes:
- Here we assume the availability of the host-side TMA API added to upstream Triton in https://github.com/triton-lang/triton/pull/4498. As of time of writing, this is not a part of the PT2 OSS Triton pin (although back-ported internally). OSS Triton pin update should be done in December 2024.
- To capture the chain of calls `t.data_ptr() --> create_{1d,2d}_tma_descriptor(ptr, ...) --> kernel[grid](tma_desc, ...)`, we add three new variable trackers: `DataPtrVariable`, `CreateTMADescriptorVariable` (for the function), `TMADescriptorVariable` (for TMA descriptor object). This is to maintain the path back from the Triton kernel to the Tensor from which the TMA descriptor has been created.
- The newly introduced variables have `reconstruct` methods used in case of graph breaks.
- The `tma_descriptor_metadata` extracted from the captured `create_{1d,2d}_tma_descriptor` calls is propagated through the HOPs in Dynamo and AOTAutograd to be used by the downstream compiler (e.g., Inductor). See the unit tests for how the captured HOP arguments look like.
- In the Dynamo-captured fx graph, we replace the TMA descriptor arguments of the Triton kernel by the underlying Tensors, to be able to track the input/output relationships in terms of Tensors.
- In the Triton kernel mutation analysis pass (in AOTAutograd), we use the `tt.experimental_descriptor_store` TTIR op to detect mutations of the underlying tensors via TMA descriptors. So that downstream AOTAutograd can perform functionalizations as required.
- JIT Inductor and AOT Inductor support will be implemented in follow-up PRs.
Differential Revision: [D64404928](https://our.internmc.facebook.com/intern/diff/D64404928)
Pull Request resolved: https://github.com/pytorch/pytorch/pull/137677
Approved by: https://github.com/zou3519
See `test_inline_closure_returned_by_another_function_and_captures` and #136814 for more context.
In #90286, we introduced an optimization so that for captured cells that are unmodified during a Dynamo trace, `UserFunctionVariable` will represent them as variable of the cell's actual value, rather than a `NewCellVariable`.
Later on we introduced more mechanisms to model such cells across function calls (#104222), and across function calls where `NestedUserFunctionVariable::bind_args` need to look up further in the parent frames (#106491) to find these cells' values.
This patch removes `InlinedClosureVariable` in favor of a simpler modelling, which is also more consistent with what was introduced in #90286, i.e., just model these cells as their contents, in `symbolic_locals`.
This fixes#136814 because resolution of `InlinedClosureVariable` to the underlying cell content value happens in
`NestedUserFunctionVariable::bind_args`, which requires Dynamo to have the value in scope at the function call site (when Dynamo does inlining), but's not always the case (as the test case shows). However, if we model the cells in `symbolic_locals`, we never need such resolution, and the values are directly stored into the `NestedUserFunctionVariable::closure` upon the function creation, at which point Dynamo always has the cell value in `symbolic_locals` for look up.
Fixes#136814.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/137510
Approved by: https://github.com/williamwen42
Partially addresses https://github.com/pytorch/pytorch/issues/128150
When you have big sums of values, we end up computing long chains of
binary addition in our FX graph representation. Not only is this ugly,
it also is quadratic, as the sympy.Add constructor is O(N) in number
of arguments. Instead, ensure that we maintain the summation as a
single FX node so we can do the entire addition all in one go.
update_hint_regression benchmark, before and after:
```
update_hint_regression,compile_time_instruction_count,2648328980
update_hint_regression,compile_time_instruction_count,2563748678
```
Signed-off-by: Edward Z. Yang <ezyang@meta.com>
Pull Request resolved: https://github.com/pytorch/pytorch/pull/136429
Approved by: https://github.com/isuruf
In #136512, we fixed handling for tl.constexpr and dynamic shapes: if a symint is passed to tl.constexpr, you should specialize on it, because tl.constexpr implies needing to know the concrete value at compile time.
However, when using triton_op, capture_triton, or non-strict export, the regression remains (and #136512 might technically regress some specific export scenarios) - see [Richard's comment](https://github.com/pytorch/pytorch/pull/136512/files#r1775999871).
This fixes these scenarios: implement the handling differently depending on whether we're expecting a SymNodeVariable or a SymInt(/SymBool/SymFloat)
Pull Request resolved: https://github.com/pytorch/pytorch/pull/136686
Approved by: https://github.com/zou3519
In #136512, we fixed handling for tl.constexpr and dynamic shapes: if a symint is passed to tl.constexpr, you should specialize on it, because tl.constexpr implies needing to know the concrete value at compile time.
However, when using triton_op, capture_triton, or non-strict export, the regression remains (and #136512 might technically regress some specific export scenarios) - see [Richard's comment](https://github.com/pytorch/pytorch/pull/136512/files#r1775999871).
This fixes these scenarios: implement the handling differently depending on whether we're expecting a SymNodeVariable or a SymInt(/SymBool/SymFloat)
Pull Request resolved: https://github.com/pytorch/pytorch/pull/136686
Approved by: https://github.com/zou3519
> Ignore FSDP2 forward hook side-effects in AC
Under AC, FSDP2 does not rely on forward hook to all-gather weights to do recomputation, instead it relies on pre-backward hook to do this job:
451eaf0ff2/torch/distributed/_composable/fsdp/_fsdp_state.py (L219-L220)
So when we use `speculate_subgraph` to trace the utils.checkpoint AC region, we don't actually need to worry about FSDP2 forward hook's side effects and can safely ignore it, because we are not and we don't expect to re-run the FSDP2 forward hook during backward recomputation.
----
Test commands:
- `pytest -rA test/distributed/_composable/fsdp/test_fully_shard_compile.py::TestFullyShardCompile::test_nested_fully_shard_backend_inductor`
- `pytest -rA test/distributed/_composable/fsdp/test_fully_shard_compile.py::TestFullyShardCompile::test_transformer_backend_inductor`
Pull Request resolved: https://github.com/pytorch/pytorch/pull/134997
Approved by: https://github.com/zou3519
ghstack dependencies: #135727
Changes:
1. Move `polyfill.py` -> `polyfills/__init__.py`. It can be used as `polyfill.xxx` -> `polyfills.xxx`.
2. Move submodule loading from `polyfills/__init__.py` to `polyfills/loader.py`.
Merge `polyfill.py` and `polyfills/` packages. Each polyfill module have its own namespace for better code organization.
The ultimate goal is make `polyfills/__init__.py` empty and all polyfill functions move to its own namespace.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/133977
Approved by: https://github.com/jansel
Add decorator `torch.compiler.substitute_in_graph` to register polyfill for unsupported C++ function to avoid graph break. This API provides an official way to add support for dynamo for third-party C extensions. Also, it can be used to simplify our implementation for `torch._dynamo.polyfill`.
5ee070266f/torch/_dynamo/variables/builtin.py (L97-L107)
Example:
```python
>>> import operator
>>> operator.indexOf([1, 2, 3, 4, 5], 3)
2
>>> torch.compile(operator.indexOf, fullgraph=True)([1, 2, 3, 4, 5], 3)
Unsupported: ...
>>> @torch.compiler.substitute_in_graph(operator.indexOf)
... def indexOf(sequence, x):
... for i, item in enumerate(sequence):
... if item is x or item == x:
... return i
... raise ValueError("sequence.index(x): x not in sequence")
>>> torch.compile(operator.indexOf, fullgraph=True)([1, 2, 3, 4, 5], 3)
2
```
Pull Request resolved: https://github.com/pytorch/pytorch/pull/133712
Approved by: https://github.com/jansel
Add decorator `torch.compiler.substitute_in_graph` to register polyfill for unsupported C++ function to avoid graph break. This API provides an official way to add support for dynamo for third-party C extensions. Also, it can be used to simplify our implementation for `torch._dynamo.polyfill`.
5ee070266f/torch/_dynamo/variables/builtin.py (L97-L107)
Example:
```python
>>> import operator
>>> operator.indexOf([1, 2, 3, 4, 5], 3)
2
>>> torch.compile(operator.indexOf, fullgraph=True)([1, 2, 3, 4, 5], 3)
Unsupported: ...
>>> @torch.compiler.substitute_in_graph(operator.indexOf)
... def indexOf(sequence, x):
... for i, item in enumerate(sequence):
... if item is x or item == x:
... return i
... raise ValueError("sequence.index(x): x not in sequence")
>>> torch.compile(operator.indexOf, fullgraph=True)([1, 2, 3, 4, 5], 3)
2
```
Pull Request resolved: https://github.com/pytorch/pytorch/pull/133712
Approved by: https://github.com/jansel
Add decorator `torch.compiler.substitute_in_graph` to register polyfill for unsupported C++ function to avoid graph break. This API provides an official way to add support for dynamo for third-party C extensions. Also, it can be used to simplify our implementation for `torch._dynamo.polyfill`.
5ee070266f/torch/_dynamo/variables/builtin.py (L97-L107)
Example:
```python
>>> import operator
>>> operator.indexOf([1, 2, 3, 4, 5], 3)
2
>>> torch.compile(operator.indexOf, fullgraph=True)([1, 2, 3, 4, 5], 3)
Unsupported: ...
>>> @torch.compiler.substitute_in_graph(operator.indexOf)
... def indexOf(sequence, x):
... for i, item in enumerate(sequence):
... if item is x or item == x:
... return i
... raise ValueError("sequence.index(x): x not in sequence")
>>> torch.compile(operator.indexOf, fullgraph=True)([1, 2, 3, 4, 5], 3)
2
```
Pull Request resolved: https://github.com/pytorch/pytorch/pull/133712
Approved by: https://github.com/jansel
I didn't test this path when creating the orchestrator. This PR fixes
that path to work in the capture_triton path. The problem is that we are
handling a value that is an int (in the capture_triton path) and a
ConstantVariable (in the Dynamo triton path) so we abstract that out in
the orchestrator.
Test Plan:
- new tests
Pull Request resolved: https://github.com/pytorch/pytorch/pull/132143
Approved by: https://github.com/oulgen
TritonKernelVariable's logic tells us how to go from a user-defined
triton kernel and a grid to a call to the triton_kernel_wrapper_mutation
HOP. We want to re-use this in a setting without Dynamo; in the next PR
up, we create a new decorator (capture_triton) that, when applied to a
triton kernel, transforms a call to the triton kernel into a call
to the triton_kernel_wrapper_mutation HOP.
Test Plan:
- existing tests
Pull Request resolved: https://github.com/pytorch/pytorch/pull/130177
Approved by: https://github.com/oulgen, https://github.com/ydwu4
Hard to write tests. This PR makes many test pass in the stack such as
`PYTORCH_TEST_WITH_DYNAMO=1 pytest test/test_ao_sparsity.py::TestComposability::test_convert_without_squash_mask`
Pull Request resolved: https://github.com/pytorch/pytorch/pull/129858
Approved by: https://github.com/mlazos
ghstack dependencies: #129830
Significant bytecode generation API change!
The new suggested convention to generating bytecode to call a function is now to wrap instructions that push a callable to the stack with `add_push_null`, then that callable is called with `create_call_function` with `push_null=False` (see diff for examples).
In Python 3.13, NULL is now expected to be pushed after the callable. In <=3.12, the NULL was pushed before the callable. This change abstracts away the exact placement of the NULL, but the developer must be aware that a NULL may be needed when codegen'ing a callable.
This abstraction also reduces the need for the `push_null=True` option in `create_call_function`, which removes the need to rotate a NULL to the right place on the stack with a sequence of `SWAP` instructions.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/129172
Approved by: https://github.com/jansel
Improve Dynamo to support the FSDP2 `use_training_state()` context manager.
Test command:
`
pytest -rA test/distributed/_composable/fsdp/test_fully_shard_compile.py::TestFullyShardCompile::test_dynamo_trace_use_training_state
`
Pull Request resolved: https://github.com/pytorch/pytorch/pull/127854
Approved by: https://github.com/yanboliang
This is a short-term fix (for 2.4). In the longer term we should
fix https://github.com/pytorch/pytorch/issues/128430
The problem is that warnings.warn that are inside Dynamo print
all the time. Python warnings are supposed to print once, unless their
cache is reset: Dynamo ends up resetting that cache everytime it runs.
As a workaround we provide our own warn_once cache that is keyed on the
warning msg. I am not worried about this increasing memory usage because
that's effectively what python's warnings.warn cache does.
Test Plan:
- fix tests.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/128456
Approved by: https://github.com/anijain2305
