[Fix] State the configured MoE-DP width in the weight-cache fingerprint

[ch-wan]

This PR is part of a stack (oldest at bottom):

• [Feature] Add --attn-dp-size and deprecate --enable-dp-attention #41818
• [Refactor] Enter draft TP scopes by attention ownership and drop ModelRunner.tp_group #41817
• [Refactor] Add make_pp_layers so models stop handing their PP position down #41816
• [Refactor] Stop passing models' layers the placement they already read #41815
• [Refactor] Let FusedMoE's weight-loading helpers read the layer's MoE-TP rank #41814
• [Refactor] Keep only the TP and PP groups on the model runner #41813
• [Refactor] Drop model placement attributes and parameters nothing reads #41812
• [Refactor] Drop placement values nothing reads #41811
• [Fix] Make the Solar model constructible and runnable #41810
• [Fix] Size the prefill delayer's gather buffer to its TP group #41809
• [Fix] Pass the draft's attention ownership to DFLASH's eager LiLiCorr scope #41808
• [Fix] Shard MoE WNA16 and Quark INT4-FP8 weights by the MoE placement #41807
• -> [Fix] State the configured MoE-DP width in the weight-cache fingerprint #41806
• [Fix] Read TensorCast's WORLD placement under its current names #41805

Motivation

The engine fills CacheConfig.moe_dp_size with get_moe_cp_size(), the width of the MoE communicator. The daemon fills it with the configured moe_dp_size. CacheConfig.matches() compares for equality. Whenever attn_cp_size > moe_dp_size, the MoE-DP group aliases the attention-CP group and is wider than configured, so every fetch from the daemon fails with Daemon config mismatch!.

Modifications

• The engine now states the configured moe_dp_size, as the daemon does. The ranks already agree: both sides read moe_dp_rank from the published placement, which takes the attention-CP rank under the same aliasing.
• test_weight_cache_daemon.py gains TestWeightCacheDaemonQwen3MoeAttnCP: Qwen3-30B-A3B served through the daemon with --attn-cp-size 2 --enable-prefill-cp --cp-strategy zigzag. The 2-GPU suite's est_time rises to cover it.

Accuracy Tests

H200:

• New test class on the parent commit: the daemon logs Config mismatch: {'moe_dp_size': (1, 2)} and the client fails to load. On this PR: 3 passed, about two minutes.
• Qwen3-30B-A3B, daemon and client with --tp 2 --attn-cp-size 2 --enable-prefill-cp --cp-strategy zigzag --cuda-graph-backend-prefill=disabled: weights load over IPC, GSM8K (200 questions) 0.910.
• test/registered/unit at this PR's head, compared with main: no new failures.

Speed Tests and Profiling

Not applicable.

Checklist

• Format your code according to the Format code with pre-commit.
• Add unit tests according to the Run and add unit tests.
• Update documentation according to Write documentations.
• Provide accuracy and speed benchmark results according to Test the accuracy and Benchmark the speed.
• Follow the SGLang code style guidance.

Review and Merge Process

1. Ping Merge Oncalls to start the process. See the PR Merge Process.
2. Get approvals from CODEOWNERS and other reviewers.
3. Trigger CI tests with comments or contact authorized users to do so.
   • Common commands include /tag-and-rerun-ci, /tag-run-ci-label, /rerun-failed-ci
4. After green CI and required approvals, ask Merge Oncalls or people with Write permission to merge the PR.

---

CI States

Latest PR Test (Base): 🚫 Run #36772507415
Latest PR Test (Extra): 🚫 Run #36772507303
Latest PR Test (AMD ROCm 10): ❌ Run #36772507201