perf(dsv4): adapt tuned AITER ASM FP8 decode

[yhl-amd]

Summary

• Adapt DeepSeek-V4 native FP8 decode to the swapped gfx950 ASM decode kernel from [ASM] [MLA v4 nm] Swap gfx950 asm decode kernel aiter#5195.

• Pass AITER's KV split plan ([ASM] [HIP] [MLA v4 nm] Split plan helper and persistent decode kernel with in-kernel split and merge aiter#6126) to the decode ASM kernel. Whoever builds a decode CSR builds its split plan next to it, in its own persistent split_indptr buffer:

  • V4 metadata builder, verify/decode: one plan per layer type (SWA window, CSA window + index_topk, HCA window + max_model_len // 128), sized for the padded row count.
  • V4 metadata builder, MTP draft (prepare_mtp_decode): SWA plan for the draft rows.
  • DSpark draft (DSparkIndexBuffers): built with the draft CSR at stage 0 and read by every stage (B*T rows, W + T KV).

  Each plan depends only on the row count and config, so CUDA graph capture and replay agree. The owner rewrites the buffer in place before each forward.

• sparse_attn_v4_paged_decode takes a single split_plan; the model selects it in the same ratio branch as kv_indices / kv_indptr.

• v4_decode_split_plan is the only place that checks for AITER's helper. With an AITER build that lacks it, plans are None and AITER keeps its own split pick.

This PR is ASM-only. It adds or changes no FlyDSL/Triton decode implementation, route, environment variable, or kernel file.

Dependencies

• [ASM] [MLA v4 nm] Swap gfx950 asm decode kernel aiter#5195 (merged): gfx950 v4 nm decode code object
• [ASM] [HIP] [MLA v4 nm] Split plan helper and persistent decode kernel with in-kernel split and merge aiter#6126: get_mla_v4_nm_split_plan. Without it this PR leaves the split pick to AITER.

Test plan

• tests/test_v4_prefill_asm_decode_dispatch.py, tests/test_paged_attention_dispatch.py, tests/test_dspark.py: pass with and without the AITER helper
• ruff / black on changed files
• DeepSeek-V4-Pro-DSpark, FP8 KV, TP8/EP8/DPA8, --cudagraph-mode FULL, real draft acceptance, with [ASM] [HIP] [MLA v4 nm] Split plan helper and persistent decode kernel with in-kernel split and merge aiter#6126: graph capture, correct outputs, DSpark acceptance unchanged vs. the stock image
• GitHub CI

🤖 Generated with Claude Code

[github-actions]

🏷️ CI Guide

Runs automatically on every eligible PR before approval:

• ✅ Pre Checkin: Black, Ruff, catalog schema validation, non-GPU unit tests

Heavy model tests:

• ✅ Run after the PR is approved and Pre Checkin passes
• ✅ Run immediately when an approval review is submitted
• ✅ Can be requested before approval with labels

Label	Tests
ci:full	Run all heavy PR model tests: native ATOM, vLLM, and SGLang
ci:atom	Run native ATOM model accuracy tests
ci:vllm	Run ATOM vLLM OOT model accuracy tests
ci:sglang	Run ATOM SGLang model accuracy tests

Heavy jobs are skipped when the PR is not approved and no matching ci:* label is present.
Add labels via the sidebar or gh pr edit 2411 --add-label <label>

[yhl-amd]

Superseded by #2462, which now carries these commits (tuned ASM FP8 decode with caller-built split plans) together with the HCA persistent-kernel path, merged with current main and with the non-GPU unit tests fixed. Closing in favour of that PR.