[GG] fix(pcie): isolate target and draft graph channels

[voipmonitor]

Summary

• bind target and speculative CUDA graph captures to independent SparkInfer DCP channels
• default PCIe oneshot all-reduce to per-stream channels
• right-size oneshot buffers from their actual dispatch cutoffs
• retain explicit capture ownership when CUDA stream identities are recycled
• keep MRV2 profiling, workspace-lane, and MoE scheduling changes out of this PR

Root cause

Target and draft graph managers may reuse a CUDA capture stream identity while owning different replayable graphs. A process-global channel mapping can therefore return a channel captured by another graph manager. Channels must follow the enclosing graph capture rather than the recycled stream key alone.

Dependency

Requires local-inference-lab/sparkinfer#59.

This branch is based directly on dev/gilded-gnosis; it is not stacked on another temporary PR branch. It contains the PCIe/channel portion split out of #149.

Validation

• tests/distributed/test_b12x_fused_all_reduce.py
• tests/distributed/test_dcp_a2a.py
• release CUDA image run: 52 passed
• Ruff lint and format checks passed
• git diff --check passed

[coderabbitai]

Warning

Review limit reached

@voipmonitor, you've reached your PR review limit, so we couldn't start this review.

Next review available in: 59 minutes

Enable usage-based reviews in Billing to review now. Otherwise, wait until the next included review is available.
You're only billed for reviews past your plan's rate limits ($0.25/file).

How can I continue?

After more reviews become available, a review can be triggered using the @coderabbitai review command as a PR comment. Alternatively, push new commits to this PR.

To avoid repeated limits, reduce automatic review volume by pausing incremental auto-reviews earlier, using label-based review opt-in, excluding WIP or generated PR titles, or requesting reviews manually when the PR is ready. If your team needs uninterrupted high-volume reviews, an organization admin can enable usage-based reviews.

How do review limits work?

CodeRabbit enforces per-developer PR review limits for each organization. Most developers receive the normal plan review availability.

For paid Pro and Pro+ PR reviews, CodeRabbit uses adaptive limits for sustained high-volume activity. When a developer's recent PR review activity reaches the 95th percentile or higher among CodeRabbit users, additional reviews become available more gradually as earlier reviews age out of the rolling window.

Please refer docs for additional details.

Review details

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro Plus

Run ID: 737299f5-d94a-4d09-bc39-4c751c59c4f0

📥 Commits

Reviewing files that changed from the base of the PR and between 73e4a8c and faa01d4.

📒 Files selected for processing (6)

• tests/distributed/test_b12x_fused_all_reduce.py
• tests/distributed/test_dcp_a2a.py
• vllm/distributed/device_communicators/custom_all_reduce.py
• vllm/distributed/parallel_state.py
• vllm/envs.py
• vllm/v1/attention/ops/dcp_alltoall.py

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch fix/gg-pcie-graph-channel-isolation-20260721

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[voipmonitor]

Folded unchanged into #149 as the second reviewable commit. #149 now carries workspace ownership, graph-channel ownership, and MRV2 profiling as one direct-GG lifecycle fix, so there is no hidden merge-order dependency. Closing this PR as superseded; no implementation or tests were dropped.