perf: tune bf16 gemv and add opt-in fused swiglu

- deepen common-shape BF16 GEMV tuning with warp-row tiling for LLaMA/Qwen2/GPT-NeoX/OPT decode projections
- add fused BF16 up/gate SwiGLU CUDA primitive with ASTRAI_SWIGLU=0/1/auto dispatch
- keep the unfused linear backend as the default path; auto enables no shape until per-architecture checkpoint gates pass
- fall back to the linear/torch chain when kernels are absent, on CPU, in training, or outside supported M/K/dtype shapes
- add gemv/swiglu benchmark scripts, dispatch and parity tests, and kernel documentation

Benchmark: NVIDIA L20 (sm_89), CUDA 12.8, PyTorch 2.11.0+cu128, idle GPU. AstrAI 1B config (24 layers, hidden 1536, vocab 100000), BF16, prompt 128, 32 greedy decode tokens, CUDA graphs enabled, A/B in separate interleaved processes (3 rounds, 8 trials each, medians). Default vs ASTRAI_SWIGLU=1 per generate call: batch 1 134.8->129.1 ms (+4.44%), batch 2 136.2->130.9 ms (+4.06%), batch 4 145.5->140.3 ms (+3.66%). Greedy output identical at batch 1, differs at batch 2/4, so auto stays unfused by default; kernelless fallback verified bit-identical greedy.

docs/developer/decode_linear_benchmark.md

@@ -20,3 +20,22 @@ change the decode batch sizes. Compare each GPU architecture only with its own
 baseline; do not use absolute A100-versus-L20 numbers as a dispatch criterion.
 Keep the raw JSON as the source of truth and generate tables with
 `--markdown-output` rather than transcribing measurements by hand.
+
+For direct A/B coverage of the custom kernel and guarded dispatcher across
+traditional LLaMA and GPT-NeoX decode shapes, use:
+
+```bash
+CUDA_VISIBLE_DEVICES=0 PYTHONPATH=. python scripts/tools/benchmark_gemv_common.py \
+  --suite all --family traditional --m 2 4 \
+  --output results/gemv_common.json
+```
+
+The kernel suite compares the directly callable primitive with `F.linear`.
+Use repeatable `--shape-label` and `--chain-label` filters for a focused run.
+The synthetic-chain suite alternates `ASTRAI_GEMV=0` and `auto`, includes
+dependent MLP work and Python dispatch, and rotates through distinct weights.
+Pass `--candidate-mode 1` to characterize a family before adding it to the
+automatic shape table; the checked-in final evidence always uses `auto`.
+It is deliberately not labeled a whole-model throughput benchmark. Both
+suites report median/p90 CUDA-event latency plus maximum absolute error,
+relative L2 error, and row-wise argmax parity.