feat: add LightningAttentionInfinilm for MiniMax linear attention

[y258dd]

Summary

• Add LightningAttentionInfinilm, an InfiniLM-specific indexed-pool lightning attention operator.

• Add the shared operator interface and metadata in src/base/lightning_attention_infinilm.h.

• Add a CPU reference implementation in src/native/cpu/ops/lightning_attention_infinilm/lightning_attention_infinilm.h.

• Add CUDA device kernel, launcher, and NVIDIA vendor binding under src/native/cuda/ and src/native/cuda/nvidia/.

• Add an Ascend ACLNN implementation in src/native/ascend/ops/lightning_attention_infinilm/kernel.h.

• Add pytest coverage in tests/test_lightning_attention_infinilm.py.

Motivation

MiniMax-style linear attention uses a recurrent lightning-attention state with a per-head decay and an indexed state pool:

ratio = exp(-slope)
state = ratio * state + outer(k_t, v_t)
out_t = q_t @ state

PyTorch has no directly corresponding operator. The closest public reference is Flash-Linear-Attention's fused_recurrent_lightning_attn, but this implementation also has the InfiniLM-specific indexed state-pool contract.

The operator is needed by the MiniMax model support work in InfiniLM.
Closes #

Type of Change

• feat — new feature / new operator / new platform
• fix — bug fix
• perf — performance improvement (no behavioral change)
• refactor — code restructuring without behavior change
• test — adding or fixing tests only
• docs — documentation only
• build / ci — build system or CI configuration
• chore — tooling, formatting, or other non-code changes
• Breaking change (requires a ! in the Conventional Commits prefix or a BREAKING CHANGE: footer)

Platforms Affected

• CPU (WITH_CPU)
• NVIDIA (WITH_NVIDIA)
• Iluvatar (WITH_ILUVATAR)
• MetaX (WITH_METAX)
• Cambricon (WITH_CAMBRICON)
• Moore (WITH_MOORE)
• Ascend (WITH_ASCEND)
• PyTorch C++ bindings (WITH_TORCH)
• Build system / CMake / CI
• Python bindings / user-facing API

Smoke Test Result

Affected-platform builds succeeded.

Operator-specific smoke commands and results:

# CPU
python3 -m pytest tests/test_lightning_attention_infinilm.py --devices cpu -v
24 passed

# NVIDIA RTX 5090, CUDA 12.8
pytest tests/test_lightning_attention_infinilm.py -v
48 passed  # CPU + CUDA

# Ascend 910B1, CANN 9.0.0, torch_npu 2.10.0
python3 -m pytest tests/test_lightning_attention_infinilm.py --devices ascend -v
24 passed```

## Test Results on Supported Platforms

<!--
Per `CONTRIBUTING.md` §Pull Requests, build and run the smoke test set on every
affected platform. Use `smoke passed`, `full passed`, or `N/A - not affected` in
the result columns. Run the full suite for high-risk changes, release prep,
maintainer spot checks, or changes affecting shared build, dispatch, wrapper
generation, or cross-platform behavior.

If an affected platform was not tested, state the reason and tag a reviewer or
owner with access. Reviewers may request full-suite or full-platform validation
when the risk profile justifies it.
-->

| Platform   | Affected | Build / Smoke Result | Full Result / Notes |
| ---------- | :------: | ------------------- | ------------------- |
| CPU        |   Yes    | build passed        | operator tests passed: 24 |
| NVIDIA     |   Yes    | build passed        | operator tests passed: 48 (24 CPU + 24 CUDA) |
| Iluvatar   |   No     | N/A                 | N/A |
| MetaX      |   No     | N/A                 | N/A |
| Cambricon  |   No     | N/A                 | N/A |
| Moore      |   No     | N/A                 | N/A |
| Ascend     |   Yes    | build passed        | operator tests passed: 24 on 910B1 / CANN 9.0.0 |
<details>
<summary>Full `pytest` output (optional)</summary>

```text
CPU:    24 passed
NVIDIA: 48 passed
Ascend: 24 passed```

</details>

## Benchmark / Performance Impact

N/A. This PR is a correctness-first implementation. The Ascend path composes ACLNN operators and does not claim fused-kernel performance; an AscendC fused kernel is a possible follow-up.
<!--
Required for `perf` PRs; optional otherwise. Describe the benchmark harness,
shapes, dtypes, hardware, and include baseline vs. new numbers. If the PR is
not performance-sensitive, write "N/A".
-->

## Notes for Reviewers
- `initial_state` is intentionally passed as a mutable `Tensor` because the destination row is updated.
- The row selected by `initial_state_indices` is never modified; the source row is copied into a workspace state first and the destination row is written only after all tokens have been processed.
- Requests are independent. A destination row must not also be the source row of another request in the same call.
- The current pytest matrix covers `int32` state indices. The operator interface implements `int64` as well, but that path is not currently parameterized in the test matrix.
- Ascend uses `aclnnMul`, `aclnnMatmul`, and `aclnnAdd` with the existing `WorkspacePool`. FP32 matmul uses `cube_math_type=0` to avoid FP32 downcast on 910B.
- The CUDA implementation assumes `head_dim` fits within one block.
<!--
Anything reviewers should focus on: subtle invariants, known trade-offs,
follow-up work intentionally left out of scope, etc.
-->

[y258dd]

Closing this PR because the source branch does not follow the required
feat/... branch naming convention in CONTRIBUTING.md.

Reopening from feat/lightning-attention-infinilm.