bench: add criterion benchmarks for protocol, bandwidth, TCP RX scan, and EC-SRP5

Adds four Criterion.rs benchmark suites to measure hot-path performance
and demonstrate the impact of Sprints 1–3 optimizations:

- benches/protocol.rs    — Command & StatusMessage serialize/deserialize
- benches/bandwidth.rs   — BandwidthState atomics, budget, interval math
- benches/tcp_rx_scan.rs — memchr SIMD scan vs naive O(n) loop (55× faster
                           on 256KB buffers with status at end)
- benches/ecsrp5.rs      — WCurve::new() heavy math vs cached LazyLock
                           (~123,000× faster access)

Also adds BENCHMARKS.md with usage instructions and example results.

Visibility changes (bench-only):
- scan_status_message is now pub (was #[cfg(test)] only)
- WCurve and WCURVE are now pub in ecsrp5.rs

dev-dependencies: criterion + pprof (optional flamegraph support)

benches/bandwidth.rs

@@ -0,0 +1,79 @@
+use criterion::{black_box, criterion_group, criterion_main, Criterion};
+use btest_rs::bandwidth::{BandwidthState, calc_send_interval, advance_next_send};
+use std::sync::atomic::Ordering;
+use std::time::{Duration, Instant};
+
+fn bench_atomic_fetch_add(c: &mut Criterion) {
+    let state = BandwidthState::new();
+    c.bench_function("bandwidth_rx_bytes_fetch_add", |b| {
+        b.iter(|| {
+            black_box(state.rx_bytes.fetch_add(1500, Ordering::Relaxed));
+        })
+    });
+    c.bench_function("bandwidth_tx_bytes_fetch_add", |b| {
+        b.iter(|| {
+            black_box(state.tx_bytes.fetch_add(32768, Ordering::Relaxed));
+        })
+    });
+}
+
+fn bench_spend_budget(c: &mut Criterion) {
+    // Unlimited budget (fast path)
+    let unlimited = BandwidthState::new();
+    c.bench_function("spend_budget_unlimited", |b| {
+        b.iter(|| black_box(unlimited.spend_budget(black_box(1500))))
+    });
+
+    // Limited budget
+    let limited = BandwidthState::new();
+    limited.byte_budget.store(1_000_000_000, Ordering::SeqCst);
+    c.bench_function("spend_budget_limited", |b| {
+        b.iter(|| black_box(limited.spend_budget(black_box(1500))))
+    });
+}
+
+fn bench_calc_send_interval(c: &mut Criterion) {
+    c.bench_function("calc_interval_100mbps_1500b", |b| {
+        b.iter(|| black_box(calc_send_interval(black_box(100_000_000), black_box(1500))))
+    });
+    c.bench_function("calc_interval_1gbps_32768b", |b| {
+        b.iter(|| black_box(calc_send_interval(black_box(1_000_000_000), black_box(32768))))
+    });
+    c.bench_function("calc_interval_unlimited", |b| {
+        b.iter(|| black_box(calc_send_interval(black_box(0), black_box(1500))))
+    });
+}
+
+fn bench_advance_next_send(c: &mut Criterion) {
+    let iv = Duration::from_micros(120);
+    let now = Instant::now();
+    let mut next = now;
+    c.bench_function("advance_next_send", |b| {
+        b.iter(|| {
+            let r = advance_next_send(&mut next, iv, now);
+            black_box(r);
+        });
+        next = now;
+    });
+}
+
+fn bench_summary(c: &mut Criterion) {
+    let state = BandwidthState::new();
+    // Pre-populate some values so loads are real
+    state.total_tx_bytes.store(1_000_000_000, Ordering::Relaxed);
+    state.total_rx_bytes.store(2_000_000_000, Ordering::Relaxed);
+    state.intervals.store(100, Ordering::Relaxed);
+    c.bench_function("bandwidth_summary", |b| {
+        b.iter(|| black_box(state.summary()))
+    });
+}
+
+criterion_group!(
+    bandwidth_benches,
+    bench_atomic_fetch_add,
+    bench_spend_budget,
+    bench_calc_send_interval,
+    bench_advance_next_send,
+    bench_summary
+);
+criterion_main!(bandwidth_benches);