Speed up get_current_state_ids

Using _simple_select_list is fairly expensive for functions that return
a lot of rows and/or get called a lot. (This is because it carefully
constructs a list of dicts).

get_current_state_ids gets called a lot on startup and e.g. when the IRC
bridge decided to send tonnes of joins/leaves (as it invalidates the
cache). We therefore replace it with a custon txn function that builds
up the final result dict without building up and intermediate
representation.

fix typo in synctl help

Don't double encode replication data

Reduce rate of USER_SYNC repl commands

Rearrange TCP replication metrics

Move to using TCP replication

Deduplicate new deviceinbox rows for replication

Advance replication streams even if nothing is listening

Otherwise the streams don't advance and steadily fall behind, so when a
worker does connect either a) they'll be streamed lots of old updates or
b) the connection will fail as the streams are too far behind.

Replace HTTP replication with TCP replication (Server side part)

Cull spurious PreserveLoggingContexts

Accept join events from all servers

Make sure that we accept join events from any server, rather than just the
origin server, to make the federation join dance work correctly.

(Fixes #1893).

In `MessageHandler`, remove `yield` on call to `Notifier.on_new_room_event`:
it doesn't return anything anyway.

Remove `PreserveLoggingContext` around calls to `Notifier.on_new_room_event`;
there is no problem if the logcontext is set when calling it.