Turn inliner into a pass and use it in rewriter & optimizer

[justinchuby]

Use passes in optimizer and rewriter.

1. By opting into using the pass infra early, we get the benefit of getting the additional features in pass infra w/o having to pay higher refactoring cost in the future. We will be able to add more sophisticated debug utilities/snapshot capabilities etc. to the passes.
2. Since we are offering the pass infra to users, we can start validating it internally by using it here. If order altering becomes a valid use case we can expect users may need that and we can create appropriate facilities to support the usage.

[Copilot]

Pull Request Overview

This PR refactors the inlining and rewriting functionality by converting them into passes that integrate into the pass manager, improving modularity and consistency in the optimizer pipeline.

• Introduced RewritePass in rewriter to encapsulate pattern-based rewriting.
• Updated the optimizer to use a sequential pass manager that includes InlinePass, FoldConstantsPass, RewritePass, and unused removal passes.
• Refactored the inlining functionality by replacing _Inliner with InlinePass, which now returns a boolean flag indicating modifications.

Reviewed Changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated 1 comment.

File	Description
onnxscript/rewriter/init.py	Added RewritePass and updated rewrite function to use pass manager.
onnxscript/optimizer/_optimizer.py	Reorganized optimizer passes to integrate new InlinePass and RewritePass.
onnxscript/optimizer/_inliner.py	Converted _Inliner to InlinePass with a revised state reset and return value.

[codecov]

❌ 35 Tests Failed:

Tests completed	Failed	Passed	Skipped
16306	35	16271	3251

View the top 3 failed test(s) by shortest run time

onnxscript.backend.onnx_export_test.TestOnnxBackEnd::test_export2python_produces_correct_onnx_script_model_0289_test_convinteger_without_padding

Stack Traces | 0.003s run time

onnxscript\backend\onnx_export_test.py:137: in extract_functions
    mod = importlib.import_module(import_name)
C:\hostedtoolcache\windows\Python\3.11.9\x64\Lib\importlib\__init__.py:126: in import_module
    return _bootstrap._gcd_import(name[level:], package, level)
E   ModuleNotFoundError: No module named 'tests.onnx_backend_test_code.test_convinteger_without_padding'

The above exception was the direct cause of the following exception:
.nox\test_torch_nightly\Lib\site-packages\parameterized\parameterized.py:620: in standalone_func
    return func(*(a + p.args), **p.kwargs, **kw)
onnxscript\backend\onnx_export_test.py:271: in test_export2python_produces_correct_onnx_script_model
    functions = extract_functions(backend_test.name, code, self.test_folder)
onnxscript\backend\onnx_export_test.py:139: in extract_functions
    raise AssertionError(
E   AssertionError: Unable to import 'tests.onnx_backend_test_code.test_convinteger_without_padding' (e=No module named 'tests.onnx_backend_test_code.test_convinteger_without_padding') (file: 'D:\\a\\onnxscript\\onnxscript\\tests\\onnx_backend_test_code\\test_convinteger_without_padding.py', absolute path: 'D:\\a\\onnxscript\\onnxscript\\tests\\onnx_backend_test_code\\test_convinteger_without_padding.py', current folder: D:\a\onnxscript\onnxscript
E   ---- CONTENT --
E   import numpy
E   from onnx import TensorProto
E   from onnx.helper import make_tensor
E   from onnxscript import script, external_tensor
E   from onnxscript.values import Opset
E   from onnxscript.onnx_types import INT32, UINT8
E   from onnxscript.onnx_opset import opset10
E   
E   @script()
E   def bck_test_convinteger_without_padding(x: UINT8[1,1,3,3], w: UINT8[1,1,2,2], x_zero_point: UINT8) -> (INT32[1,1,2,2]):
E       y = opset10.ConvInteger(x, w, x_zero_point)
E       return y

onnxscript.backend.onnx_export_test.TestOnnxBackEnd::test_export2python_produces_correct_onnx_script_model_0630_test_max_float16

Stack Traces | 0.003s run time

onnxscript\backend\onnx_export_test.py:137: in extract_functions
    mod = importlib.import_module(import_name)
C:\hostedtoolcache\windows\Python\3.11.9\x64\Lib\importlib\__init__.py:126: in import_module
    return _bootstrap._gcd_import(name[level:], package, level)
E   ModuleNotFoundError: No module named 'tests.onnx_backend_test_code.test_max_float16'

The above exception was the direct cause of the following exception:
.nox\test_ort_nightly\Lib\site-packages\parameterized\parameterized.py:620: in standalone_func
    return func(*(a + p.args), **p.kwargs, **kw)
onnxscript\backend\onnx_export_test.py:271: in test_export2python_produces_correct_onnx_script_model
    functions = extract_functions(backend_test.name, code, self.test_folder)
onnxscript\backend\onnx_export_test.py:139: in extract_functions
    raise AssertionError(
E   AssertionError: Unable to import 'tests.onnx_backend_test_code.test_max_float16' (e=No module named 'tests.onnx_backend_test_code.test_max_float16') (file: 'D:\\a\\onnxscript\\onnxscript\\tests\\onnx_backend_test_code\\test_max_float16.py', absolute path: 'D:\\a\\onnxscript\\onnxscript\\tests\\onnx_backend_test_code\\test_max_float16.py', current folder: D:\a\onnxscript\onnxscript
E   ---- CONTENT --
E   import numpy
E   from onnx import TensorProto
E   from onnx.helper import make_tensor
E   from onnxscript import script, external_tensor
E   from onnxscript.values import Opset
E   from onnxscript.onnx_types import FLOAT16
E   from onnxscript.onnx_opset import opset13
E   
E   @script()
E   def bck_test_max_float16(data_0: FLOAT16[3], data_1: FLOAT16[3]) -> (FLOAT16[3]):
E       result = opset13.Max(data_0, data_1)
E       return result

onnxscript.backend.onnx_export_test.TestOnnxBackEnd::test_export2python_produces_correct_onnx_script_model_0681_test_mod_uint32

Stack Traces | 0.003s run time

onnxscript\backend\onnx_export_test.py:137: in extract_functions
    mod = importlib.import_module(import_name)
C:\hostedtoolcache\windows\Python\3.11.9\x64\Lib\importlib\__init__.py:126: in import_module
    return _bootstrap._gcd_import(name[level:], package, level)
E   ModuleNotFoundError: No module named 'tests.onnx_backend_test_code.test_mod_uint32'

The above exception was the direct cause of the following exception:
.nox\test_torch_nightly\Lib\site-packages\parameterized\parameterized.py:620: in standalone_func
    return func(*(a + p.args), **p.kwargs, **kw)
onnxscript\backend\onnx_export_test.py:271: in test_export2python_produces_correct_onnx_script_model
    functions = extract_functions(backend_test.name, code, self.test_folder)
onnxscript\backend\onnx_export_test.py:139: in extract_functions
    raise AssertionError(
E   AssertionError: Unable to import 'tests.onnx_backend_test_code.test_mod_uint32' (e=No module named 'tests.onnx_backend_test_code.test_mod_uint32') (file: 'D:\\a\\onnxscript\\onnxscript\\tests\\onnx_backend_test_code\\test_mod_uint32.py', absolute path: 'D:\\a\\onnxscript\\onnxscript\\tests\\onnx_backend_test_code\\test_mod_uint32.py', current folder: D:\a\onnxscript\onnxscript
E   ---- CONTENT --
E   import numpy
E   from onnx import TensorProto
E   from onnx.helper import make_tensor
E   from onnxscript import script, external_tensor
E   from onnxscript.values import Opset
E   from onnxscript.onnx_types import UINT32
E   from onnxscript.onnx_opset import opset13
E   
E   @script()
E   def bck_test_mod_uint32(x: UINT32[3], y: UINT32[3]) -> (UINT32[3]):
E       z = opset13.Mod(x, y)
E       return z

To view more test analytics, go to the Test Analytics Dashboard
_{📋 Got 3 mins? Take this short survey to help us improve Test Analytics.}

[gramalingam]

I think this is causing a regression when no rewrite-rules are specified. I think we should retain an if pattern_rewrite_rules condition and avoid the RewritePass if no rules are specified.