re_match_program: one whole-program build per worktree at a time (a lock file); concurrent runs collided

harbor/reverse.py

@@ -598,15 +598,19 @@ def match_program(config, targets=None, decomp=None):
                                                text=True, encoding="utf-8", errors="replace", timeout=timeout,
                                                cwd=str(Path(script).parent),
                                                creationflags=getattr(subprocess, "CREATE_NO_WINDOW", 0))
-    built = run(["build", *extra], 1800)
+    # one build per worktree at a time: two runs on the same checkout write the same build folder
-    if built.returncode != 0:
+    lock_dir = Path(config.get("debugger_temp") or tempfile.gettempdir())
-        raise HarborError("the whole-program build failed: " + (built.stdout + built.stderr).strip()[-800:])
+    lock = lock_dir / ("ltcg-" + re.sub(r"[^A-Za-z0-9]+", "_", str(decomp or "main")) + ".lock")
-    with tempfile.TemporaryDirectory(dir=config.get("debugger_temp")) as tmp:
+    with _exclusive(lock, wait_seconds=1800):
-        out = Path(tmp) / "check.json"
+        built = run(["build", *extra], 1800)
-        checked = run(["check", *extra, "--json", str(out)] + (["--only", ",".join(wanted)] if wanted else []), 900)
+        if built.returncode != 0:
-        if not out.exists():
+            raise HarborError("the whole-program build failed: " + (built.stdout + built.stderr).strip()[-800:])
-            raise HarborError("the whole-program check failed: " + (checked.stdout + checked.stderr).strip()[-800:])
+        with tempfile.TemporaryDirectory(dir=config.get("debugger_temp")) as tmp:
-        rows = json.loads(out.read_text(encoding="utf-8"))
+            out = Path(tmp) / "check.json"
+            checked = run(["check", *extra, "--json", str(out)] + (["--only", ",".join(wanted)] if wanted else []), 900)
+            if not out.exists():
+                raise HarborError("the whole-program check failed: " + (checked.stdout + checked.stderr).strip()[-800:])
+            rows = json.loads(out.read_text(encoding="utf-8"))
     counts = {}
     for r in rows:
         counts[r.get("status")] = counts.get(r.get("status"), 0) + 1
@@ -642,3 +646,36 @@ def behaviour_check(config, targets, decomp=None, precision=24):
         rows = json.loads(out.read_text(encoding="utf-8"))
     keep = ("address", "name", "verdict", "reason", "weak", "counts", "notes")
     return [{k: r[k] for k in keep if k in r} for r in rows]
+
+
+class _exclusive:
+    """A lock file shared between processes: wait for it, hold it, remove it (stale after two hours)."""
+
+    def __init__(self, path, wait_seconds=1800):
+        self.path, self.wait = Path(path), wait_seconds
+
+    def __enter__(self):
+        import time
+        deadline = time.monotonic() + self.wait
+        while True:
+            try:
+                fd = os.open(str(self.path), os.O_CREAT | os.O_EXCL | os.O_WRONLY)
+                os.write(fd, str(os.getpid()).encode())
+                os.close(fd)
+                return self
+            except FileExistsError:
+                try:
+                    if time.time() - self.path.stat().st_mtime > 7200:
+                        self.path.unlink()
+                        continue
+                except FileNotFoundError:
+                    continue
+                if time.monotonic() > deadline:
+                    raise HarborError(f"another whole-program build on this worktree is still running ({self.path.name}); try again later") from None
+                time.sleep(5)
+
+    def __exit__(self, *exc):
+        try:
+            self.path.unlink()
+        except FileNotFoundError:
+            pass