review: fix review findings (correctness, hardening, doc-accuracy)

First recursive code-review-harness pass over the authored SDK code (sim/,
config-lint, flowgen, drivers/{relays,freshness}, build/, ci.yml). Four parallel
dimension reviewers; all findings at every severity corrected per workspace rule.

Correctness / reliability:
- freshness.c min_budget_ms: use a `seen` flag, not `best==0`, as the empty
  sentinel — a zero-tolerance (max_stale_ms==0) binding was silently widened to a
  looser neighbour's budget. Regression test added; still 66/66 MC/DC.
- config-lint parse_reserved_ranges: match `reg` as a whole property token (ident
  boundary before, `=` after) so `reg-names` / a `region-*` label no longer
  mis-parses into a bogus reserved range.
- config-lint loader check: fail closed — flag any loader with a LOAD_ADDR that is
  not a known-safe boot component, instead of only known MCU names, so a future
  coprocessor ("Rtos"/"Bl32") can't slip past the 0x40000-brick gate.
- build-kernel.sh: sha256 verification is now mandatory (refuse to build if the pin
  is missing) and the mktemp scratch tree is removed on exit (trap), while a
  caller-provided WORK is left intact for CI artifact upload.

Test quality:
- freshness: added the age==max_stale boundary case and a clock-wraparound
  (now < last_ok) fail-safe-to-UNKNOWN test.
- relays: unsetenv(WARDEN_GPIO_ROOT) at main() so the NULL-env arm is hermetic.

Security / CI:
- ci.yml: top-level `permissions: contents: read` (badges overrides to write);
  pin taiki-e/install-action to commit SHA (v2.86.7).

Maintainability / docs:
- drivers/enforce-mcdc.sh: one shared, name-derived gate replaces the two
  copy-pasted per-driver scripts; Makefiles call ../../enforce-mcdc.sh.
- docs/architecture.md: §3/§4/§6/§7 rewritten to match reality — NPU/RGA models,
  config-lint, and the relays+freshness MC/DC harnesses are done; kernel §6 now
  reflects the 5.10->6.18.46 forward-port (ADR-0001), not the superseded plan44/6.6.
- README: status blurb + layout table corrected (kernel/, .github/; stale ci/ and
  patches/ descriptions fixed). hpmcu "8 tests" -> 7 in docs.
- freshness.{c,h}: ADR reference points at flare-edge ADR-0004 (warden-sdk's
  ADR-0004 is the CI runner — number collision).
- normalize rustfmt drift across sim/ + config-lint.

All green: sim 37 tests, config-lint 8 tests, both drivers 100% MC/DC (relays
40/40, freshness 66/66), clippy clean under -D warnings, gitleaks clean.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_017wB8KB3MMQztRDXCMCkPrf

drivers/freshness/test/test_freshness.c

@@ -44,6 +44,9 @@ static void test_decide(void) {
     EXPECT(warden_fresh_decide(FRESH_UNKNOWN, false, false, 0,   100) == FRESH_RENDER_UNKNOWN);
     EXPECT(warden_fresh_decide(FRESH_UNKNOWN, true,  false, 200, 100) == FRESH_RENDER_UNKNOWN);
     EXPECT(warden_fresh_decide(FRESH_UNKNOWN, true,  false, 50,  100) == FRESH_RENDER_NOCHANGE);
+    /* boundary: age == max_stale is NOT stale (guards a `>`->`>=` regression that
+     * MC/DC alone would not catch — both outcomes are already covered above). */
+    EXPECT(warden_fresh_decide(FRESH_UNKNOWN, true,  false, 100, 100) == FRESH_RENDER_NOCHANGE);
 }
 
 /* --- 2. bind: !produce, !render, valid, and table-full (FRESH_MAX=2) --- */
@@ -121,10 +124,18 @@ static void test_invalidate(void) {
     EXPECT(g_render_calls == 1);
 }
 
-/* --- 6. min_budget: best==0 first, then max_stale<best true and false --- */
+/* --- 6. min_budget: unseen first, then max_stale<best true and false --- */
 static void test_min_budget(void) {
     warden_fresh_reset();
     EXPECT(warden_fresh_min_budget_ms() == 0);     /* nothing bound */
+    /* regression (correctness review): a zero-tolerance binding (max_stale=0) must
+     * win the minimum, not be mistaken for the "nothing scanned" sentinel and
+     * widened to a looser neighbour's budget. */
+    warden_fresh_bind((void*)1, fake_produce, NULL, 0,   "s", fake_render, (void*)9);
+    warden_fresh_bind((void*)1, fake_produce, NULL, 300, "s", fake_render, (void*)9);
+    EXPECT(warden_fresh_min_budget_ms() == 0);
+    warden_fresh_reset();
+    EXPECT(warden_fresh_min_budget_ms() == 0);     /* nothing bound (post-reset) */
     warden_fresh_bind((void*)1, fake_produce, NULL, 300, "s", fake_render, (void*)9); /* best=0->300 */
     warden_fresh_bind((void*)1, fake_produce, NULL, 100, "s", fake_render, (void*)9); /* 100<300 -> 100 */
     EXPECT(warden_fresh_min_budget_ms() == 100);
@@ -160,6 +171,22 @@ static void test_false_arms(void) {
     EXPECT(g_render_calls == 0);
 }
 
+/* --- 8. clock wraparound: a uint32 ms counter wraps every ~49.7 days, so a tick
+ *        can arrive with now_ms < last_ok_ms. The unsigned `now - last_ok_ms` then
+ *        underflows to a huge age; the contract must fail SAFE to UNKNOWN, never
+ *        assert the last value as if still fresh. --- */
+static void test_wraparound(void) {
+    warden_fresh_reset();
+    warden_fresh_bind((void*)1, fake_produce, NULL, 100, "s", fake_render, (void*)9);
+    set_produce(FRESH_OK, "9"); reset_render();
+    warden_fresh_tick(1000);                       /* last_ok_ms = 1000, ever_ok */
+    EXPECT(g_render_calls == 1 && g_render_what == FRESH_RENDER_VALUE);
+    /* now wraps behind last_ok_ms: age underflows -> treated as stale -> UNKNOWN */
+    set_produce(FRESH_UNKNOWN, NULL); reset_render();
+    warden_fresh_tick(10);
+    EXPECT(g_render_calls == 1 && g_render_what == FRESH_RENDER_UNKNOWN);
+}
+
 int main(void) {
     test_decide();
     test_bind();
@@ -168,6 +195,7 @@ int main(void) {
     test_invalidate();
     test_min_budget();
     test_false_arms();
+    test_wraparound();
     fprintf(stderr, "%d checks, %d failures\n", g_checks, g_fail);
     return g_fail ? 1 : 0;
 }