Test that TORCH_FEATURE_VERSION guards are used where needed

[bar-qodo]

User description

Splits each torch library registration in the 2.10 folder into its own file -- I had a script that parsed kernel.cpp to do this but I felt like forcing this responsibility on the user might be less error prone

Compiles each file targetting 2.9 and asserts that compilation fails. (There are 2 2.9 kernels we use as negative tests where compilation is expected to succeed)

Stack from ghstack (oldest at bottom):

• Add sum and sum_out to stable ops pytorch/pytorch#168062
• Move MemoryFormat/Layout to headeronly pytorch/pytorch#168034
• -> Test that TORCH_FEATURE_VERSION guards are used where needed pytorch/pytorch#167962
• Test libtorch_agnostic with TORCH_TARGET_VERSION on target pytorch version pytorch/pytorch#167804
• Split libtorch agnostic tests by feature version pytorch/pytorch#167803
• Fix TORCH_FEATURE_VERSION guards pytorch/pytorch#167802
• Follow up on #161891 move additions to stable shim and use version guards pytorch/pytorch#168025

---

PR Type

Tests, Enhancement

---

Description

• Split monolithic kernel.cpp into individual function files for better modularity

• Created comprehensive test suite verifying TORCH_FEATURE_VERSION guards for 2.10+ APIs

• Added negative tests (mv_tensor_accessor files) to validate test infrastructure

• Dynamically generates test methods for each .cpp and .cu file in csrc directory

---

Diagram Walkthrough

flowchart LR
  A["kernel.cpp<br/>monolithic file"] -->|"split into"| B["Individual function files<br/>my__foreach_mul.cpp<br/>test_device_*.cpp<br/>etc."]
  B -->|"tested by"| C["test_version_compatibility.py<br/>compiles with 2.9.0"]
  C -->|"expects failure"| D["2.10+ specific code<br/>fails compilation"]
  C -->|"expects success"| E["Negative tests<br/>mv_tensor_accessor_*."]

Loading

File Walkthrough

	Relevant files
Tests	3 files test_version_compatibility.py New test suite for version compatibility verification        +300/-0  mv_tensor_accessor_cpu.cpp Added negative test file for CPU tensor accessor                  +40/-0    mv_tensor_accessor_cuda.cu Added negative test file for CUDA tensor accessor                +47/-0
Refactoring	16 files kernel.cpp Removed monolithic kernel file split into separate files  +0/-205  my__foreach_mul.cpp Extracted foreach_mul function into dedicated file              +20/-0    my__foreach_mul_.cpp Extracted foreach_mul_ function into dedicated file            +19/-0    make_tensor_clones_and_call_foreach.cpp Extracted make_tensor_clones_and_call_foreach function into dedicated file +41/-0    my_empty.cpp Extracted my_empty function into dedicated file                    +25/-0    my_reshape.cpp Extracted my_reshape function into dedicated file                +17/-0    my_view.cpp Extracted my_view function into dedicated file                      +20/-0    test_tensor_device.cpp Extracted test_tensor_device function into dedicated file +17/-0    test_device_constructor.cpp Extracted test_device_constructor function into dedicated file +37/-0    test_device_equality.cpp Extracted test_device_equality function into dedicated file +14/-0    test_device_set_index.cpp Extracted test_device_set_index function into dedicated file +17/-0    test_device_index.cpp Extracted test_device_index function into dedicated file  +14/-0    test_device_is_cuda.cpp Extracted test_device_is_cuda function into dedicated file +14/-0    test_device_is_cpu.cpp Extracted test_device_is_cpu function into dedicated file +14/-0    test_parallel_for.cpp Extracted test_parallel_for function into dedicated file  +49/-0    test_get_num_threads.cpp Extracted test_get_num_threads function into dedicated file +14/-0
Enhancement	1 files tensor_accessor_kernel.h Added shared tensor accessor kernel header file                    +28/-0

---

[qodo-merge-pro]

PR Compliance Guide 🔍

Below is a summary of compliance checks for this PR:

Security Compliance
🟢	No security concerns identified No security vulnerabilities detected by AI analysis. Human verification advised for critical code.
Ticket Compliance
⚪	🎫 No ticket provided Create ticket/issue
Codebase Duplication Compliance
⚪	Codebase context is not defined Follow the guide to enable codebase context checks.
Custom Compliance
🟢	Generic: Meaningful Naming and Self-Documenting Code Objective: Ensure all identifiers clearly express their purpose and intent, making code self-documenting Status: Passed Learn more about managing compliance generic rules or creating your own custom rules
	Generic: Secure Error Handling Objective: To prevent the leakage of sensitive system information through error messages while providing sufficient detail for internal debugging. Status: Passed Learn more about managing compliance generic rules or creating your own custom rules
	Generic: Secure Logging Practices Objective: To ensure logs are useful for debugging and auditing without exposing sensitive information like PII, PHI, or cardholder data. Status: Passed Learn more about managing compliance generic rules or creating your own custom rules
	Generic: Security-First Input Validation and Data Handling Objective: Ensure all data inputs are validated, sanitized, and handled securely to prevent vulnerabilities Status: Passed Learn more about managing compliance generic rules or creating your own custom rules
⚪	Generic: Comprehensive Audit Trails Objective: To create a detailed and reliable record of critical system actions for security analysis and compliance. Status: No audit logs: The tests and helper compilation routines perform actions without any auditing or logging of critical actions, but as test code this may be acceptable and not subject to audit requirements. Referred Code import os import subprocess import tempfile from pathlib import Path from torch.testing._internal.common_utils import run_tests, TestCase from torch.utils.cpp_extension import CUDA_HOME, include_paths as torch_include_paths class FunctionVersionCompatibilityTest(TestCase): """Test that all function files require PyTorch 2.10+.""" @classmethod def setUpClass(cls): """Set up test environment once for all tests.""" cls.csrc_dir = Path(__file__).parent / "libtorch_agnostic_2_10" / "csrc" cls.build_dir = Path(tempfile.mkdtemp(prefix="version_check_")) cls.pytorch_includes = [ f"-I{path}" for path in torch_include_paths(device_type="cpu") ] ... (clipped 194 lines) Learn more about managing compliance generic rules or creating your own custom rules
	Generic: Robust Error Handling and Edge Case Management Objective: Ensure comprehensive error handling that provides meaningful context and graceful degradation Status: Limited error handling: Subprocess compilation errors are only asserted on success/failure with optional printing under an env var and fixed timeout, without retries or granular handling of edge cases like missing compilers or include paths which may be acceptable for tests. Referred Code Compile a C++ file with TORCH_TARGET_VERSION=2.9.0. Returns (success, error_message). """ torch_version_2_9 = "0x0209000000000000" cmd = [ "g++", "-c", "-std=c++17", f"-DTORCH_TARGET_VERSION={torch_version_2_9}", f"-I{source_file.parent}", # For includes in same directory *self.pytorch_includes, ] # Add CUDA flags if available if self.cuda_available: cmd.extend(self.cuda_includes) cmd.extend([str(source_file), "-o", str(output_file)]) result = subprocess.run(cmd, capture_output=True, text=True, timeout=30) ... (clipped 61 lines) Learn more about managing compliance generic rules or creating your own custom rules

Compliance status legend

🟢 - Fully Compliant
🟡 - Partial Compliant
🔴 - Not Compliant
⚪ - Requires Further Human Verification
🏷️ - Compliance label

[qodo-merge-pro]

PR Code Suggestions ✨

Explore these optional code suggestions:

Category	Suggestion	Impact
Possible issue	Fix function signature in schema Update the schema for mv_tensor_accessor_cpu to match its function signature by removing the extra Tensor res argument. test/cpp_extensions/libtorch_agnostic_2_10_extension/libtorch_agnostic_2_10/csrc/mv_tensor_accessor_cpu.cpp [34-36] STABLE_TORCH_LIBRARY_FRAGMENT(libtorch_agnostic_2_10, m) { - m.def("mv_tensor_accessor_cpu(Tensor res, Tensor m, Tensor v) -> Tensor"); + m.def("mv_tensor_accessor_cpu(Tensor m, Tensor v) -> Tensor"); } Apply / Chat Suggestion importance[1-10]: 9 __ Why: The suggestion correctly identifies a critical bug where the function signature of mv_tensor_accessor_cpu mismatches its registered schema, which would cause a compilation failure.	High
More