Fix FP16 NEON build on AArch64 CPUs without FP16FML support

[Harish-endee]

1. FP16 NEON distance functions (L2Sqr, InnerProductSim) rely on vfmlalq_low_f16 and vfmlalq_high_f16 intrinsics. These intrinsics require the AArch64 FP16FML extension. Not all AArch64 CPUs support FP16FML, leading to compilation failures on unsupported targets.

2. Added compatibility fallback implementations

• Fallback uses universally supported NEON instructions: vcvt_f32_f16 and vfmaq_f32
• Introduced compile-time macro dispatch to select the appropriate implementation
• CPUs with FP16FML continue using the optimized single-instruction path with no overhead
• Fallback code is only compiled when the FP16FML extension is unavailable

[github-actions]

VectorDB Benchmark - Ready To Run

CI Passed ([lint + unit tests] (https://github.com/endee-io/endee/actions/runs/24066965981)) - benchmark options unlocked.

Post one of the command below. Only members with write access can trigger runs.

---

Available Modes

Mode	Command	What runs
Dense	/correctness_benchmarking dense	HNSW insert throughput · query P50/P95/P99 · recall@10 · concurrent QPS
Hybrid	/correctness_benchmarking hybrid	Dense + sparse BM25 fusion · same suite + fusion latency overhead

---

Infrastructure

Server	Role	Instance
Endee Server	Endee VectorDB — code from this branch	t2.large
Benchmark Server	Benchmark runner	t3a.large

Both servers start on demand and are always terminated after the run — pass or fail.

---

How Correctness Benchmarking Works

1. Post /correctness_benchmarking <mode>
2. Endee Server Create  →  this branch's code deployed  →  Endee starts in chosen mode
3. Benchmark Server Create  →  benchmark suite transferred
4. Benchmark Server runs correctness benchmarking against Endee Server
5. Results posted back here  →  pass/fail + full metrics table
6. Both servers terminated   →  always, even on failure

After a new push, CI must pass again before this menu reappears.