Skipping VirtualMemoryResource tests on WSL

[rparolin]

Skipping VirtualMemoryResource tests on WSL

When running the VirtualMemoryResources tests under the Windows Subsystem for Linux (WSL) allocation requests throw the following error:

> raise CUDAError(f"{name.decode()}: {expl}")
E cuda.core.experimental._utils.cuda_utils.CUDAError: CUDA_ERROR_INVALID_DEVICE: This indicates that the device ordinal supplied by the user does not correspond to a valid CUDA device or that the action requested is invalid for the specified device.
cuda/core/experimental/_utils/cuda_utils.pyx:78: CUDAError

See known limitations for WSL2.
https://docs.nvidia.com/cuda/wsl-user-guide/index.html#known-limitations-for-linux-cuda-applications

[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.

[rparolin]

/ok to test 46af80d

[github-actions]

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

[leofang]

@rparolin in what ways VMM raises an error on WSL? Could you show a full traceback to what you see?

Asking because we were told that VMM is supported by the Windows WDDM driver mode (#968 (comment)), and WSL is a pass-through of WDDM according to @ksimpson-work, and so by deduction it should be supported. This is a bit surprising... 🤔

[keenan-simpson]

This is not the right place to be commenting, but skipping tests on windows because something isn't supported in TCC feels heavy handed. a utility function to check if the system is in TCC mode seems more appropriate. RE: VMM I would also be curious to know what the issues on WSL are for VMM, if this isn't already fully resolved