Reduce cache size by not storing deferreds

Currently the cache descriptors store deferreds rather than raw values,
this is a simple way of triggering only one database hit and sharing the
result if two callers attempt to get the same value.

However, there are a few caches that simply store a mapping from string
to string (or int). These caches can have a large number of entries,
under the assumption that each entry is small. However, the size of a
deferred (specifically the size of ObservableDeferred) is signigicantly
larger than that of the raw value, 2kb vs 32b.

This PR therefore changes the cache descriptors to store the raw values
rather than the deferreds.

As a side effect cached storage function now either return a deferred or
the actual value, as the cached list decriptor already does. This is
fine as we always end up just yield'ing on the returned value
eventually, which handles that case correctly.

synapse/storage/receipts.py

@@ -47,10 +47,13 @@ class ReceiptsStore(SQLBaseStore):
         # Returns an ObservableDeferred
         res = self.get_users_with_read_receipts_in_room.cache.get((room_id,), None)
 
-        if res and res.called and user_id in res.result:
-            # We'd only be adding to the set, so no point invalidating if the
-            # user is already there
-            return
+        if res:
+            if isinstance(res, defer.Deferred) and res.called:
+                res = res.result
+            if user_id in res:
+                # We'd only be adding to the set, so no point invalidating if the
+                # user is already there
+                return
 
         self.get_users_with_read_receipts_in_room.invalidate((room_id,))