dcp: don't fail boot when B12X DCP collectives don't support the world size

[voipmonitor]

Problem

The B12X PCIe DCP channel exists only for world sizes 2/4/8, and the runtime dispatchers (dcp_b12x_all_gather_heads, dcp_a2a_lse_reduce) already fall back to NCCL collectives per call. The dedicated DCP warmup added in "Optimize B12X DCP collectives and warmup" turned that graceful fallback into a boot failure: with TP6 + DCP3/DCP6 (GLM head66 configs that booted fine on the v13 stack) warmup_b12x_dcp_a2a raises

RuntimeError: B12X PCIe DCP query all-gather is unavailable for the configured attention geometry

and EngineCore dies. This regression is on the current base branch / v14 image — it is independent of PR #78.

Fix

• warmup_b12x_dcp_a2a: log once and skip when the DCP world size is not 2/4/8 (still raises for genuinely broken geometries at supported sizes). This also covers the DSv4 module path in kernel_warmup.py.
• MLAAttention: don't set dcp_b12x for unsupported DCP sizes, so neither warmup nor the per-step dispatch attempts the PCIe channel at all.

Verification

GLM-5.2 NVFP4 TP6 on the v14 image (vllm02f5b41-b12xe44cb77) with the v13 memory shape (GMU 0.957, max_model_len 128000, max_num_batched_tokens 2048), MTP3, MOE_MODE=a16:

Mode	KV cache tokens	short-ctx tok/s	3k-ctx tok/s	CJK
TP6 DCP1	168,384	131.1	106.7	0
TP6 DCP6 (fixed)	989,125	74.9	74.5	0

Without the fix, TP6 DCP6 dies at warmup with the RuntimeError above. Head66 virtual-TP padding (attention heads 64 -> 66) and PCIe allreduce at world size 6 both work on this stack.

New unit test: test_warmup_skips_unsupported_world_size (no dist/GPU needed).

Note: the same commit was originally pushed as the second commit of #78; it is split out here because it fixes a pre-existing base-branch regression and can merge independently. #78 will be rebased to contain only the hybrid-dispatch feature.

🤖 Generated with Claude Code

[coderabbitai]

Warning

Review limit reached

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

Next review available in: 33 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: be4d500d-35e0-44ee-81ee-df50d9acc47f

📥 Commits

Reviewing files that changed from the base of the PR and between 872145f and 59944f4.

📒 Files selected for processing (3)

• tests/distributed/test_dcp_a2a.py
• vllm/model_executor/layers/attention/mla_attention.py
• vllm/v1/attention/ops/dcp_alltoall.py

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch fable/dcp-warmup-ws-guard-20260707

---

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.}