Minor testing rework.

[Andy-Jost]

Follow-on changes to #1176 based on review feedback.

[copy-pr-bot]

This pull request requires additional validation before any workflows can run on NVIDIA's runners.

Pull request vetters can view their responsibilities here.

Contributors can view more details about this message here.

[greptile-apps]

Greptile Overview

Greptile Summary

This PR refactors test helper infrastructure following review feedback from PR #1176. The changes consolidate platform detection logic and libc loading into the shared cuda_python_test_helpers package, replacing repetitive conditional imports across test modules with centralized utilities. The cuda_core/tests/helpers/__init__.py now uses wildcard imports to automatically expose all test helper symbols (IS_WSL, IS_WINDOWS, supports_ipc_mempool, libc) without manual re-exports. The buffers.py module removes inline platform-specific ctypes CDLL loading in favor of importing the shared libc object. Additionally, the latch kernel's nanosleep interval is reduced from 10ms to 1ms to improve test responsiveness. These changes align with the broader testing infrastructure improvements for IPC-enabled events, promoting DRY principles and reducing maintenance burden.

Confidence score: 1/5

• This PR has critical bugs that will cause immediate test failures if merged
• Score reflects a critical initialization order bug in latch.py line 63 (accessing self.busy_wait_flag[0] before self.buffer.handle and self.timeout_cycles are fully initialized), incorrect libc loading that hardcodes libc.so.6 (breaking macOS/non-Linux platforms), and inefficient conditional logic in supports_ipc_mempool that calls _detect_wsl() instead of using the already-computed IS_WSL constant
• Pay extremely close attention to cuda_core/tests/helpers/latch.py lines 61-65 (initialization order), cuda_python_test_helpers/cuda_python_test_helpers/__init__.py lines 31-36 (platform-specific libc loading), and line 47 (redundant WSL detection)

Additional Comments (1)

1. cuda_core/tests/helpers/latch.py, line 63 (link)

   logic: accessing busy_wait_flag property before timeout_cycles is set will cause AttributeError

   Move line 65 above line63to ensure timeout_cycles exists before any property access.

_{4 files reviewed, 1 comment}

_{Edit Code Review Agent Settings | Greptile}

[Andy-Jost]

This seems like a good place to use import *, since we are trying to merge modules defined in different places, we control both, and the source has an __all__ list.

[Andy-Jost]

/ok to test

[copy-pr-bot]

/ok to test

@Andy-Jost, there was an error processing your request: E1

See the following link for more information: https://docs.gha-runners.nvidia.com/cpr/e/1/

[Andy-Jost]

/ok to test 5cb98ae

[github-actions]

Doc Preview CI
Preview removed because the pull request was closed or merged.