memory leaks

starpilot/controls/lib/curve_speed_controller.py

@@ -25,12 +25,55 @@ class CurveSpeedController:
     self.training_timer = 0
 
     curvature_data = self.starpilot_planner.params.get("CurvatureData")
-    self.curvature_data = curvature_data if isinstance(curvature_data, dict) else {}
+    self.curvature_data = self._normalize_curvature_data(curvature_data)
 
     self.required_curvatures = [str(round(road_curvature, ROUNDING_PRECISION)) for road_curvature in np.arange(MIN_CURVATURE, MAX_CURVATURE + STEP, STEP)]
 
     self.update_lateral_acceleration()
 
+  @staticmethod
+  def _bucket_curvature(road_curvature):
+    clipped_curvature = float(np.clip(road_curvature, MIN_CURVATURE, MAX_CURVATURE))
+    bucket_index = round((clipped_curvature - MIN_CURVATURE) / STEP)
+    bucketed_curvature = MIN_CURVATURE + (bucket_index * STEP)
+    return str(round(bucketed_curvature, ROUNDING_PRECISION))
+
+  @classmethod
+  def _normalize_curvature_data(cls, curvature_data):
+    if not isinstance(curvature_data, dict):
+      return {}
+
+    normalized = {}
+    for key, value in curvature_data.items():
+      if not isinstance(value, dict):
+        continue
+
+      try:
+        raw_curvature = abs(float(key))
+        average = float(value["average"])
+        count = int(value["count"])
+      except (KeyError, TypeError, ValueError):
+        continue
+
+      if count <= 0:
+        continue
+
+      bucket = cls._bucket_curvature(raw_curvature)
+      if bucket in normalized:
+        existing = normalized[bucket]
+        total_count = existing["count"] + count
+        normalized[bucket] = {
+          "average": ((existing["average"] * existing["count"]) + (average * count)) / total_count,
+          "count": total_count,
+        }
+      else:
+        normalized[bucket] = {
+          "average": average,
+          "count": count,
+        }
+
+    return normalized
+
   def log_data(self, v_ego, sm):
     self.enable_training = v_ego > CRUISING_SPEED
     self.enable_training &= not self.starpilot_planner.tracking_lead
@@ -41,9 +84,9 @@ class CurveSpeedController:
 
       if self.training_timer >= PLANNER_TIME and self.starpilot_planner.driving_in_curve and not (sm["carState"].leftBlinker or sm["carState"].rightBlinker):
         lateral_acceleration = abs(self.starpilot_planner.lateral_acceleration)
-        road_curvature = abs(round(self.starpilot_planner.road_curvature, ROUNDING_PRECISION))
+        road_curvature = self._bucket_curvature(abs(self.starpilot_planner.road_curvature))
 
-        key = str(road_curvature)
+        key = road_curvature
         if key in self.curvature_data:
           data = self.curvature_data[key]
 

starpilot/controls/lib/speed_limit_controller.py

@@ -70,11 +70,16 @@ class SpeedLimitController:
     self.previous_target = self.starpilot_planner.params.get_float("PreviousSpeedLimit")
 
     self.executor = ThreadPoolExecutor(max_workers=1)
+    self.mapbox_future = None
 
     self.session = requests.Session()
     self.session.headers.update({"Accept-Language": "en"})
     self.session.headers.update({"User-Agent": "starpilot-mapbox-speed-limit-retriever/1.0 (https://github.com/FrogAi/StarPilot)"})
 
+  def shutdown(self):
+    self.executor.shutdown(wait=False, cancel_futures=True)
+    self.session.close()
+
   @property
   def experimental_mode(self):
     return self.target == 0 and bool(getattr(self.starpilot_toggles, "slc_fallback_experimental_mode", False))
@@ -113,11 +118,9 @@ class SpeedLimitController:
       return
 
     def make_request():
+      successful = False
+      response_data = None
       try:
-        self.calling_mapbox = True
-
-        successful = False
-
         if not is_url_pingable(self.mapbox_host):
           self.segment_distance = 1000
           return None
@@ -160,8 +163,7 @@ class SpeedLimitController:
         response.raise_for_status()
 
         successful = True
-
-        return response.json()
+        response_data = response.json()
       except Exception as exception:
         print(f"Unexpected error in Mapbox request: {exception}")
       finally:
@@ -170,8 +172,7 @@ class SpeedLimitController:
         if not successful:
           self.mapbox_limit = 0
           self.segment_distance = v_ego
-
-          return None
+      return response_data
 
     def complete_request(future):
       try:
@@ -212,8 +213,18 @@ class SpeedLimitController:
         print(f"Mapbox Callback Error: {exception}")
         self.mapbox_limit = 0
         self.segment_distance = v_ego
+      finally:
+        self.mapbox_future = None
 
-    future = self.executor.submit(make_request)
+    self.calling_mapbox = True
+    try:
+      future = self.executor.submit(make_request)
+    except RuntimeError:
+      self.calling_mapbox = False
+      self.segment_distance = v_ego
+      return
+
+    self.mapbox_future = future
     future.add_done_callback(complete_request)
 
   def handle_limit_change(self, desired_source, desired_target, sm):

starpilot/controls/starpilot_planner.py

@@ -55,7 +55,7 @@ class StarPilotPlanner:
     self.tracking_lead_filter = FirstOrderFilter(0, 0.5, DT_MDL)
 
   def shutdown(self):
-    self.starpilot_vcruise.slc.executor.shutdown(wait=False, cancel_futures=True)
+    self.starpilot_vcruise.slc.shutdown()
     self.starpilot_weather.executor.shutdown(wait=False, cancel_futures=True)
 
   def update(self, now, time_validated, sm, starpilot_toggles):

starpilot/system/speed_limit_filler.py

@@ -19,9 +19,12 @@ NetworkType = log.DeviceState.NetworkType
 
 BOUNDING_BOX_RADIUS_DEGREE = 0.1
 MAX_ENTRIES = 1_000_000
+MAX_PENDING_ADDITIONS = 2_000
 MAX_OVERPASS_DATA_BYTES = 1_073_741_824
 MAX_OVERPASS_REQUESTS = 10_000
 METERS_PER_DEG_LAT = 111_320
+PENDING_FLUSH_BATCH_SIZE = 1_000
+PENDING_FLUSH_INTERVAL_S = 60.0
 VETTING_INTERVAL_DAYS = 7
 ACTIVE_SEGMENT_BUFFER_METERS = 25
 MAX_BEARING_DELTA = 45
@@ -40,8 +43,9 @@ class MapSpeedLogger:
 
     self.cached_segments = {}
 
-    self.dataset_additions = deque(maxlen=MAX_ENTRIES)
+    self.dataset_additions = deque(maxlen=MAX_PENDING_ADDITIONS)
     self.filtered_dataset = []
+    self.last_dataset_flush = time.monotonic()
 
     self.overpass_requests = self.params.get("OverpassRequests")
     self.overpass_requests.setdefault("day", datetime.now(timezone.utc).day)
@@ -80,6 +84,8 @@ class MapSpeedLogger:
 
       entry_copy = item.copy()
       entry_copy.pop("last_vetted", None)
+      if "last_vetted" in item:
+        entry_copy = {key: entry_copy[key] for key in required if key != "last_vetted"}
 
       key = json.dumps(entry_copy, sort_keys=True)
       cleaned_data[key] = item
@@ -118,13 +124,68 @@ class MapSpeedLogger:
     if filtered_dataset is None:
       filtered_dataset = self.params.get("SpeedLimitsFiltered") or []
 
-    live_match_entries = [entry for entry in filtered_dataset if self.has_live_match_fields(entry)]
-    if live_match_entries:
-      self.filtered_dataset = live_match_entries
-      return
+    if filtered_dataset and not any(self.has_live_match_fields(entry) for entry in filtered_dataset):
+      dataset = self.cleanup_dataset(self.params.get("SpeedLimits"))
+      filtered_dataset = self.enrich_filtered_dataset(dataset, filtered_dataset)
 
-    fallback_dataset = self.params.get("SpeedLimits") or []
-    self.filtered_dataset = list(self.cleanup_dataset(fallback_dataset))
+    self.filtered_dataset = [entry for entry in filtered_dataset if self.has_live_match_fields(entry)]
+
+  @staticmethod
+  def get_entry_identity(entry):
+    start_coordinates = entry.get("start_coordinates")
+    if not isinstance(start_coordinates, dict):
+      return None
+
+    try:
+      return (
+        bool(entry.get("incorrect_limit")),
+        round(float(start_coordinates["latitude"]), 6),
+        round(float(start_coordinates["longitude"]), 6),
+        str(entry.get("source", "")),
+        round(float(entry.get("speed_limit", 0)), 3),
+      )
+    except (KeyError, TypeError, ValueError):
+      return None
+
+  def enrich_filtered_dataset(self, dataset, filtered_dataset):
+    dataset_by_identity = {}
+    for entry in dataset:
+      if not self.has_live_match_fields(entry):
+        continue
+
+      identity = self.get_entry_identity(entry)
+      if identity is None or identity in dataset_by_identity:
+        continue
+
+      dataset_by_identity[identity] = entry
+
+    if not dataset_by_identity:
+      return filtered_dataset
+
+    filtered_entries = list(filtered_dataset)
+    updated = False
+    for index, entry in enumerate(filtered_entries):
+      if self.has_live_match_fields(entry):
+        continue
+
+      identity = self.get_entry_identity(entry)
+      matching_entry = dataset_by_identity.get(identity)
+      if matching_entry is None:
+        continue
+
+      filtered_entries[index] = {
+        **entry,
+        "bearing": matching_entry["bearing"],
+        "end_coordinates": matching_entry["end_coordinates"],
+        "road_name": matching_entry["road_name"],
+        "road_width": matching_entry["road_width"],
+      }
+      updated = True
+
+    if not updated:
+      return filtered_dataset
+
+    return self.cleanup_dataset(filtered_entries)
 
   def get_speed_limit_source(self):
     vision_speed_limit = self.params_memory.get_float("VisionSpeedLimit") if self.params.get_bool("VisionSpeedLimitDetection") else 0
@@ -171,6 +232,28 @@ class MapSpeedLogger:
     self.params.put("SpeedLimits", list(dataset))
     self.params.put("SpeedLimitsFiltered", list(filtered_dataset))
 
+  def flush_pending_dataset_additions(self, force=False):
+    if not self.dataset_additions:
+      if force:
+        self.last_dataset_flush = time.monotonic()
+      return
+
+    now = time.monotonic()
+    should_flush = force
+    should_flush |= len(self.dataset_additions) >= PENDING_FLUSH_BATCH_SIZE
+    should_flush |= (now - self.last_dataset_flush) >= PENDING_FLUSH_INTERVAL_S
+    if not should_flush:
+      return
+
+    existing_dataset = self.params.get("SpeedLimits") or []
+    existing_dataset.extend(self.dataset_additions)
+
+    new_dataset = self.cleanup_dataset(existing_dataset)
+    self.params.put("SpeedLimits", list(new_dataset))
+
+    self.dataset_additions.clear()
+    self.last_dataset_flush = now
+
   def find_current_speed_limit_entry(self, latitude, longitude, road_name, current_bearing):
     best_match = None
     best_score = None
@@ -371,6 +454,7 @@ class MapSpeedLogger:
       "speed_limit": speed_limit,
       "start_coordinates": self.previous_coordinates,
     })
+    self.flush_pending_dataset_additions()
 
     self.previous_coordinates = {"latitude": current_latitude, "longitude": current_longitude}
 
@@ -469,6 +553,7 @@ class MapSpeedLogger:
 
     dataset = self.cleanup_dataset(self.params.get("SpeedLimits"))
     filtered_dataset = self.cleanup_dataset(self.params.get("SpeedLimitsFiltered"))
+    filtered_dataset = self.enrich_filtered_dataset(dataset, filtered_dataset)
 
     filtered_dataset = self.vet_entries(filtered_dataset)
     self.update_params(dataset, filtered_dataset)
@@ -547,16 +632,11 @@ def main():
 
       previously_started = True
     elif previously_started:
-      existing_dataset = logger.params.get("SpeedLimits") or []
-      existing_dataset.extend(logger.dataset_additions)
-
-      new_dataset = logger.cleanup_dataset(existing_dataset)
-      logger.params.put("SpeedLimits", list(new_dataset))
+      logger.flush_pending_dataset_additions(force=True)
 
       if logger.sm["deviceState"].networkType in (NetworkType.ethernet, NetworkType.wifi):
         logger.params_memory.put_bool("UpdateSpeedLimits", True)
 
-      logger.dataset_additions.clear()
       logger.clear_live_speed_limits()
 
       previously_started = False