test: readiness signals replace sleep-based timing (N4)

- signal tests use a marker-file readiness signal: the child's command
  is 'echo ready > <marker>; exec sleep 60', the test polls
  wait_for_file(marker, 5s) — marker exists = the shell exec'd, so the
  signal lands on the real target regardless of machine load. Applied
  in tests/pty.rs (both signal tests), tests/pipe.rs (SIGTERM), and
  the src/local unit tests; wait_for_file lives in tests/common.
- cancel-cleanup post-action sleeps became bounded polls for the
  child's death (kill(pid,0) -> ESRCH, 5s deadline) — faster and
  flake-proof in both directions.
- resize/cat-stdin tests need no readiness signal at all: the adapter's
  input pump processes chunks in order — the sleeps there were pure
  latency (integration suites now ~40ms, was 200-270ms).

tests/common/mod.rs

@@ -16,6 +16,7 @@
 
 use std::collections::HashMap;
 use std::sync::Arc;
+use std::time::Duration;
 
 use alkcall::core::auth::Identity;
 use alktty::adapter::drive_session;
@@ -297,6 +298,24 @@ pub fn nanos_seed() -> u64 {
     use std::time::{SystemTime, UNIX_EPOCH};
     SystemTime::now()
         .duration_since(UNIX_EPOCH)
-        .map(|d| d.as_nanos() as u64)
-        .unwrap_or(0)
+        .expect("clock went backwards")
+        .as_nanos() as u64
+}
+
+/// Poll until `path` exists (or `timeout` elapses), sleeping 10 ms
+/// between probes. This is the readiness signal for tests that need
+/// the child to have reached a state (e.g. `exec`'d) before acting on
+/// it — replaces fixed pre-action sleeps, which fail spuriously on a
+/// loaded machine (review #001 N4). Returns `false` on timeout.
+pub async fn wait_for_file(path: &std::path::Path, timeout: Duration) -> bool {
+    let deadline = tokio::time::Instant::now() + timeout;
+    loop {
+        if path.exists() {
+            return true;
+        }
+        if tokio::time::Instant::now() >= deadline {
+            return false;
+        }
+        tokio::time::sleep(Duration::from_millis(10)).await;
+    }
 }