Skip to content

test(cuda.core): skip the VMM RDMA policy test on WSL - #3478

Merged
Andy-Jost merged 2 commits into
NVIDIA:mainfrom
Andy-Jost:ajost/wsl-rdma-test-skip
Nov 17, 2026
Merged

Andy-Jost merged 2 commits into
NVIDIA:mainfrom
Andy-Jost:ajost/wsl-rdma-test-skip

Conversation

@Andy-Jost

Copy link
Copy Markdown
Contributor

On WSL with an R580-family driver, test_vmm_allocator_policy_configuration fails with cuda.core 1.3.0: the RDMA-enabled VMM allocation raises CUDA_ERROR_UNKNOWN (tracked internally as NVBug 7103528). The test and its capability checks have not changed since 1.2.1.

GPUDirect RDMA is not supported on WSL. With a newer driver on WSL, the device reports no RDMA support and the test skips; I verified this today on driver 596.71 with cuda.core 1.2.1 and 1.3.0. The cause of the failure on the older driver is not confirmed.

This PR only stops the test from running on WSL. It does not change the library or the capability checks.

🤖 Generated with Claude Code

GPUDirect RDMA is not supported on WSL, but some drivers report the
capability there, and the RDMA-enabled VMM allocation in
test_vmm_allocator_policy_configuration then fails. Skip the test on
WSL; the capability checks stay as they are.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
@Andy-Jost Andy-Jost added this to the cuda.core 1.3.1 milestone Nov 16, 2026
@Andy-Jost Andy-Jost added P1 Medium priority - Should do test Improvements or additions to tests cuda.core Everything related to the cuda.core module to-be-backported Trigger the bot to raise a backport PR upon merge labels Nov 16, 2026
@Andy-Jost Andy-Jost self-assigned this Nov 16, 2026
@Andy-Jost
Andy-Jost requested a review from rwgk November 16, 2026 21:33
@coderabbitai

coderabbitai Bot commented Nov 16, 2026 •

Copy link
Copy Markdown

Explain Change →

No actionable comments were generated in the recent review.

Summary

Summary by CodeRabbit

  • Tests
    • The VMM allocation policy test is skipped on WSL, where reported GPUDirect RDMA support may not reflect platform compatibility. Other platforms continue to run the existing capability check and test.

The VMM allocator policy configuration test now skips on WSL. On other platforms, it continues to the existing device-capability check.

Changes: VMM test platform guard

Layer / File(s) Summary
WSL skip condition
cuda_core/tests/test_memory.py
The test module imports IS_WSL and skips the VMM allocator policy configuration test on WSL. Other platforms continue to the existing device-capability check.

Suggested reviewers: mdboom

Change: Other

Merge Risk: Minimal · up to e1435

This test-only change skips one VMM allocator policy test on WSL, where GPUDirect RDMA is unsupported. Library behavior is unchanged. No merge-blocking risk was found.

Architecture risk: Low · up to e1435

The change affects 1 system.

Changed systems: cuda_core

Architecture concerns
No architecture-level concerns identified.

Review details

Systems and components

  • observed — cuda_core (service) was modified; 1 changed file maps to changed impact.

Before / after behavior

  • observed — Modified behavior in cuda_core/tests/test_memory.py: The test module now imports IS_WSL in addition to IS_WINDOWS.
  • observed — Modified behavior in cuda_core/tests/test_memory.py: The VMM allocation policy configuration test now skips on WSL with a reason stating that GPUDirect RDMA is unsupported there even when some drivers report it as supported.
- [ ] Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts
Recent review info

Run configuration

Commits

Reviewing files that changed from the base of the PR and between 1c88ea5 and e143527.

Files selected for processing: 1

  • cuda_core/tests/test_memory.py

Included review availability: This review used your included allowance. Your plan provides up to 12 included reviews per hour; 11 remain after this review.

@rwgk

rwgk commented Nov 16, 2026

Copy link
Copy Markdown
Contributor

(Unfortunately the CI for #3359 is stuck.)

@github-actions

github-actions Bot commented Nov 17, 2026 •

Copy link
Copy Markdown
Contributor
Doc Preview CI
Preview removed because the pull request was closed or merged.

@Andy-Jost
Andy-Jost enabled auto-merge (squash) November 17, 2026 00:16
@Andy-Jost
Andy-Jost merged commit 36bc530 into NVIDIA:main Nov 17, 2026
121 checks passed
@Andy-Jost
Andy-Jost deleted the ajost/wsl-rdma-test-skip branch November 17, 2026 02:32
Andy-Jost added a commit to Andy-Jost/cuda-python that referenced this pull request Nov 17, 2026
GPUDirect RDMA is not supported on WSL, but some drivers report the
capability there, and the RDMA-enabled VMM allocation in
test_vmm_allocator_policy_configuration then fails. Skip the test on
WSL; the capability checks stay as they are.

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
(cherry picked from commit 36bc530)
github-actions Bot pushed a commit that referenced this pull request Nov 17, 2026
Removed preview folders for the following PRs:
- PR #3359
- PR #3446
- PR #3448
- PR #3450
- PR #3464
- PR #3478
- PR #3479
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

cuda.core Everything related to the cuda.core module P1 Medium priority - Should do test Improvements or additions to tests to-be-backported Trigger the bot to raise a backport PR upon merge

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants