[None][feat] Support V2 Mamba disaggregated serving

[VALLIS-NERIA]

@coderabbitai summary

Description

This PR adds disaggregated context-to-generation state transfer for the V2 Mamba hybrid cache manager.

The changes:

• Represent a Mamba layer group in transfer metadata with its state-pool address, byte stride, and layer count, and preserve those fields through serialization.
• Extract V2 convolution and SSM pages from the actual KVCMv2 physical pools, including coalesced equal-size state pools, and transfer them through the Python/NIXL transceiver while preserving legacy layer-major registration.
• Reconstruct peer-side Mamba state pages for both matching and mismatched tensor-parallel layouts.
• Select the V2 manager only after resolving the effective backend and runtime: V2 requires NIXL + Python; auto/default, explicit C++, and UCX routes remain on the C++ manager unless model runtime resolution selects Python/NIXL.
• Add model-level preferred-runtime selection for the supported hybrid architectures.
• Cover runtime/backend routing, metadata round trips, page extraction, same-TP transfer, TP-mismatch transfer, and inflight-cancellation gating.

This PR depends on #16471. It does not depend on the save-last snapshot-reuse PR. Because an upstream PR cannot use a fork-only branch as its base, this draft temporarily shows the prerequisite changes until the core PR merges and this branch is rebased.

Test Coverage

• CPU routing, transceiver, serialization, and pointer tests (64 passed)
• CUDA V2 page-table tests (2 passed)
• V2-to-V2 Python/NIXL transfer with matching TP (1 passed)
• V2-to-V2 Python/NIXL transfer with TP mismatch (1 passed)
• Mixed/V2 native transfer and physical-pool registration regressions (10 passed)
• Effective backend/runtime routing matrix (8 passed)
• pre-commit run --from-ref origin/main --to-ref HEAD

PR Checklist

• Please check this after reviewing the above items as appropriate for this PR.

[VALLIS-NERIA]

/bot run

[tensorrt-cicd]

PR_Github #59617 [ run ] triggered by Bot. Commit: d71c1ee Link to invocation

[tensorrt-cicd]

PR_Github #59617 [ run ] completed with state SUCCESS. Commit: d71c1ee
/LLM/main/L0_MergeRequest_PR pipeline #48054 completed with status: 'FAILURE'

CI Report

⚠️ Action Required:

• Please check the failed tests and fix your PR
• If you cannot view the failures, ask the CI triggerer to share details
• Once fixed, request an NVIDIA team member to trigger CI again

CI Agent Failure Analysis

Link to invocation

[VALLIS-NERIA]

/bot run

[tensorrt-cicd]

PR_Github #59671 [ run ] triggered by Bot. Commit: 9dac8ed Link to invocation

[tensorrt-cicd]

PR_Github #59671 [ run ] completed with state FAILURE. Commit: 9dac8ed
/LLM/main/L0_MergeRequest_PR pipeline #48104 completed with status: 'FAILURE'

CI Report

⚠️ Action Required:

• Please check the failed tests and fix your PR
• If you cannot view the failures, ask the CI triggerer to share details
• Once fixed, request an NVIDIA team member to trigger CI again

CI Agent Failure Analysis

Link to invocation

[VALLIS-NERIA]

Reopened as a stacked draft at VALLIS-NERIA#9 so it can target the V2 Mamba core branch and show only its single squashed commit. Closing this upstream draft as superseded.