Merge pull request #6343 from matrix-org/rav/event_auth/4

Refactor _update_auth_events_and_context_for_auth

changelog.d/6343.misc

@@ -0,0 +1 @@
+Refactor some code in the event authentication path for clarity.

synapse/handlers/federation.py

@@ -2040,8 +2040,10 @@ class FederationHandler(BaseHandler):
             auth_events (dict[(str, str)->synapse.events.EventBase]):
                 Map from (event_type, state_key) to event
 
-                What we expect the event's auth_events to be, based on the event's
+                Normally, our calculated auth_events based on the state of the room
-                position in the dag. I think? maybe??
+                at the event's position in the DAG, though occasionally (eg if the
+                event is an outlier), may be the auth events claimed by the remote
+                server.
 
                 Also NB that this function adds entries to it.
         Returns:
@@ -2091,30 +2093,35 @@ class FederationHandler(BaseHandler):
             origin (str):
             event (synapse.events.EventBase):
             context (synapse.events.snapshot.EventContext):
+
             auth_events (dict[(str, str)->synapse.events.EventBase]):
+                Map from (event_type, state_key) to event
+
+                Normally, our calculated auth_events based on the state of the room
+                at the event's position in the DAG, though occasionally (eg if the
+                event is an outlier), may be the auth events claimed by the remote
+                server.
+
+                Also NB that this function adds entries to it.
 
         Returns:
             defer.Deferred[EventContext]: updated context
         """
         event_auth_events = set(event.auth_event_ids())
 
-        if event.is_state():
+        # missing_auth is the set of the event's auth_events which we don't yet have
-            event_key = (event.type, event.state_key)
+        # in auth_events.
-        else:
-            event_key = None
-
-        # if the event's auth_events refers to events which are not in our
-        # calculated auth_events, we need to fetch those events from somewhere.
-        #
-        # we start by fetching them from the store, and then try calling /event_auth/.
         missing_auth = event_auth_events.difference(
             e.event_id for e in auth_events.values()
         )
 
+        # if we have missing events, we need to fetch those events from somewhere.
+        #
+        # we start by checking if they are in the store, and then try calling /event_auth/.
         if missing_auth:
             # TODO: can we use store.have_seen_events here instead?
             have_events = yield self.store.get_seen_events_with_rejections(missing_auth)
-            logger.debug("Got events %s from store", have_events)
+            logger.debug("Found events %s in the store", have_events)
             missing_auth.difference_update(have_events.keys())
         else:
             have_events = {}
@@ -2169,15 +2176,17 @@ class FederationHandler(BaseHandler):
                     event.auth_event_ids()
                 )
             except Exception:
-                # FIXME:
                 logger.exception("Failed to get auth chain")
 
         if event.internal_metadata.is_outlier():
+            # XXX: given that, for an outlier, we'll be working with the
+            # event's *claimed* auth events rather than those we calculated:
+            # (a) is there any point in this test, since different_auth below will
+            # obviously be empty
+            # (b) alternatively, why don't we do it earlier?
             logger.info("Skipping auth_event fetch for outlier")
             return context
 
-        # FIXME: Assumes we have and stored all the state for all the
-        # prev_events
         different_auth = event_auth_events.difference(
             e.event_id for e in auth_events.values()
         )
@@ -2191,27 +2200,22 @@ class FederationHandler(BaseHandler):
             different_auth,
         )
 
+        # now we state-resolve between our own idea of the auth events, and the remote's
+        # idea of them.
+
         room_version = yield self.store.get_room_version(event.room_id)
+        different_event_ids = [
+            d for d in different_auth if d in have_events and not have_events[d]
+        ]
 
-        different_events = yield make_deferred_yieldable(
+        if different_event_ids:
-            defer.gatherResults(
+            # XXX: currently this checks for redactions but I'm not convinced that is
-                [
+            # necessary?
-                    run_in_background(
+            different_events = yield self.store.get_events_as_list(different_event_ids)
-                        self.store.get_event, d, allow_none=True, allow_rejected=False
-                    )
-                    for d in different_auth
-                    if d in have_events and not have_events[d]
-                ],
-                consumeErrors=True,
-            )
-        ).addErrback(unwrapFirstError)
 
-        if different_events:
             local_view = dict(auth_events)
             remote_view = dict(auth_events)
-            remote_view.update(
+            remote_view.update({(d.type, d.state_key): d for d in different_events})
-                {(d.type, d.state_key): d for d in different_events if d}
-            )
 
             new_state = yield self.state_handler.resolve_events(
                 room_version,
@@ -2231,13 +2235,13 @@ class FederationHandler(BaseHandler):
             auth_events.update(new_state)
 
             context = yield self._update_context_for_auth_events(
-                event, context, auth_events, event_key
+                event, context, auth_events
             )
 
         return context
 
     @defer.inlineCallbacks
-    def _update_context_for_auth_events(self, event, context, auth_events, event_key):
+    def _update_context_for_auth_events(self, event, context, auth_events):
         """Update the state_ids in an event context after auth event resolution,
         storing the changes as a new state group.
 
@@ -2246,18 +2250,21 @@ class FederationHandler(BaseHandler):
 
             context (synapse.events.snapshot.EventContext): initial event context
 
-            auth_events (dict[(str, str)->str]): Events to update in the event
+            auth_events (dict[(str, str)->EventBase]): Events to update in the event
                 context.
 
-            event_key ((str, str)): (type, state_key) for the current event.
-                this will not be included in the current_state in the context.
-
         Returns:
             Deferred[EventContext]: new event context
         """
+        # exclude the state key of the new event from the current_state in the context.
+        if event.is_state():
+            event_key = (event.type, event.state_key)
+        else:
+            event_key = None
         state_updates = {
             k: a.event_id for k, a in iteritems(auth_events) if k != event_key
         }
+
         current_state_ids = yield context.get_current_state_ids(self.store)
         current_state_ids = dict(current_state_ids)