Newsfile

The logic is "designed" to "handle" the case where the servers view of
the state at an event doesn't match what the remote server set as the
auth events. With some hand waving the server would try and come to some
sort of conclusion of which side was correct, involving state
resolution, but this could come up with interesting results.

The entire process is unspecced and buggy, so let's just remove it.

INSTALL.md

@@ -1,10 +1,12 @@
 - [Choosing your server name](#choosing-your-server-name)
+- [Picking a database engine](#picking-a-database-engine)
 - [Installing Synapse](#installing-synapse)
   - [Installing from source](#installing-from-source)
     - [Platform-Specific Instructions](#platform-specific-instructions)
   - [Prebuilt packages](#prebuilt-packages)
 - [Setting up Synapse](#setting-up-synapse)
   - [TLS certificates](#tls-certificates)
+  - [Client Well-Known URI](#client-well-known-uri)
   - [Email](#email)
   - [Registering a user](#registering-a-user)
   - [Setting up a TURN server](#setting-up-a-turn-server)
@@ -27,6 +29,25 @@ that your email address is probably `user@example.com` rather than
 `user@email.example.com`) - but doing so may require more advanced setup: see
 [Setting up Federation](docs/federate.md).
 
+# Picking a database engine
+
+Synapse offers two database engines:
+ * [PostgreSQL](https://www.postgresql.org)
+ * [SQLite](https://sqlite.org/)
+
+Almost all installations should opt to use PostgreSQL. Advantages include:
+
+* significant performance improvements due to the superior threading and
+  caching model, smarter query optimiser
+* allowing the DB to be run on separate hardware
+
+For information on how to install and use PostgreSQL, please see
+[docs/postgres.md](docs/postgres.md)
+
+By default Synapse uses SQLite and in doing so trades performance for convenience.
+SQLite is only recommended in Synapse for testing purposes or for servers with
+light workloads.
+
 # Installing Synapse
 
 ## Installing from source
@@ -234,9 +255,9 @@ for a number of platforms.
 
 There is an offical synapse image available at
 https://hub.docker.com/r/matrixdotorg/synapse which can be used with
-the docker-compose file available at [contrib/docker](contrib/docker). Further information on
-this including configuration options is available in the README on
-hub.docker.com.
+the docker-compose file available at [contrib/docker](contrib/docker). Further
+information on this including configuration options is available in the README
+on hub.docker.com.
 
 Alternatively, Andreas Peters (previously Silvio Fricke) has contributed a
 Dockerfile to automate a synapse server in a single Docker image, at
@@ -244,7 +265,8 @@ https://hub.docker.com/r/avhost/docker-matrix/tags/
 
 Slavi Pantaleev has created an Ansible playbook,
 which installs the offical Docker image of Matrix Synapse
-along with many other Matrix-related services (Postgres database, riot-web, coturn, mxisd, SSL support, etc.).
+along with many other Matrix-related services (Postgres database, Element, coturn,
+ma1sd, SSL support, etc.).
 For more details, see
 https://github.com/spantaleev/matrix-docker-ansible-deploy
 
@@ -277,22 +299,27 @@ The fingerprint of the repository signing key (as shown by `gpg
 /usr/share/keyrings/matrix-org-archive-keyring.gpg`) is
 `AAF9AE843A7584B5A3E4CD2BCF45A512DE2DA058`.
 
-#### Downstream Debian/Ubuntu packages
+#### Downstream Debian packages
 
-For `buster` and `sid`, Synapse is available in the Debian repositories and
-it should be possible to install it with simply:
+We do not recommend using the packages from the default Debian `buster`
+repository at this time, as they are old and suffer from known security
+vulnerabilities. You can install the latest version of Synapse from
+[our repository](#matrixorg-packages) or from `buster-backports`. Please
+see the [Debian documentation](https://backports.debian.org/Instructions/)
+for information on how to use backports.
+
+If you are using Debian `sid` or testing, Synapse is available in the default
+repositories and it should be possible to install it simply with:
 
 ```
 sudo apt install matrix-synapse
 ```
 
-There is also a version of `matrix-synapse` in `stretch-backports`. Please see
-the [Debian documentation on
-backports](https://backports.debian.org/Instructions/) for information on how
-to use them.
+#### Downstream Ubuntu packages
 
-We do not recommend using the packages in downstream Ubuntu at this time, as
-they are old and suffer from known security vulnerabilities.
+We do not recommend using the packages in the default Ubuntu repository
+at this time, as they are old and suffer from known security vulnerabilities.
+The latest version of Synapse can be installed from [our repository](#matrixorg-packages).
 
 ### Fedora
 
@@ -419,6 +446,60 @@ so, you will need to edit `homeserver.yaml`, as follows:
 For a more detailed guide to configuring your server for federation, see
 [federate.md](docs/federate.md).
 
+## Client Well-Known URI
+
+Setting up the client Well-Known URI is optional but if you set it up, it will
+allow users to enter their full username (e.g. `@user:<server_name>`) into clients
+which support well-known lookup to automatically configure the homeserver and
+identity server URLs. This is useful so that users don't have to memorize or think
+about the actual homeserver URL you are using.
+
+The URL `https://<server_name>/.well-known/matrix/client` should return JSON in
+the following format.
+
+```
+{
+  "m.homeserver": {
+    "base_url": "https://<matrix.example.com>"
+  }
+}
+```
+
+It can optionally contain identity server information as well.
+
+```
+{
+  "m.homeserver": {
+    "base_url": "https://<matrix.example.com>"
+  },
+  "m.identity_server": {
+    "base_url": "https://<identity.example.com>"
+  }
+}
+```
+
+To work in browser based clients, the file must be served with the appropriate
+Cross-Origin Resource Sharing (CORS) headers. A recommended value would be
+`Access-Control-Allow-Origin: *` which would allow all browser based clients to
+view it.
+
+In nginx this would be something like:
+```
+location /.well-known/matrix/client {
+    return 200 '{"m.homeserver": {"base_url": "https://<matrix.example.com>"}}';
+    add_header Content-Type application/json;
+    add_header Access-Control-Allow-Origin *;
+}
+```
+
+You should also ensure the `public_baseurl` option in `homeserver.yaml` is set
+correctly. `public_baseurl` should be set to the URL that clients will use to
+connect to your server. This is the same URL you put for the `m.homeserver`
+`base_url` above.
+
+```
+public_baseurl: "https://<matrix.example.com>"
+```
 
 ## Email
 
@@ -437,7 +518,7 @@ email will be disabled.
 
 ## Registering a user
 
-The easiest way to create a new user is to do so from a client like [Riot](https://riot.im).
+The easiest way to create a new user is to do so from a client like [Element](https://element.io/).
 
 Alternatively you can do so from the command line if you have installed via pip.
 

README.rst

@@ -45,7 +45,7 @@ which handle:
 - Eventually-consistent cryptographically secure synchronisation of room
   state across a global open network of federated servers and services
 - Sending and receiving extensible messages in a room with (optional)
-  end-to-end encryption[1]
+  end-to-end encryption
 - Inviting, joining, leaving, kicking, banning room members
 - Managing user accounts (registration, login, logout)
 - Using 3rd Party IDs (3PIDs) such as email addresses, phone numbers,
@@ -82,9 +82,6 @@ at the `Matrix spec <https://matrix.org/docs/spec>`_, and experiment with the
 
 Thanks for using Matrix!
 
-[1] End-to-end encryption is currently in beta: `blog post <https://matrix.org/blog/2016/11/21/matrixs-olm-end-to-end-encryption-security-assessment-released-and-implemented-cross-platform-on-riot-at-last>`_.
-
-
 Support
 =======
 
@@ -115,12 +112,11 @@ Unless you are running a test instance of Synapse on your local machine, in
 general, you will need to enable TLS support before you can successfully
 connect from a client: see `<INSTALL.md#tls-certificates>`_.
 
-An easy way to get started is to login or register via Riot at
-https://riot.im/app/#/login or https://riot.im/app/#/register respectively.
+An easy way to get started is to login or register via Element at
+https://app.element.io/#/login or https://app.element.io/#/register respectively.
 You will need to change the server you are logging into from ``matrix.org``
 and instead specify a Homeserver URL of ``https://<server_name>:8448``
 (or just ``https://<server_name>`` if you are using a reverse proxy).
-(Leave the identity server as the default - see `Identity servers`_.)
 If you prefer to use another client, refer to our
 `client breakdown <https://matrix.org/docs/projects/clients-matrix>`_.
 
@@ -137,7 +133,7 @@ it, specify ``enable_registration: true`` in ``homeserver.yaml``. (It is then
 recommended to also set up CAPTCHA - see `<docs/CAPTCHA_SETUP.md>`_.)
 
 Once ``enable_registration`` is set to ``true``, it is possible to register a
-user via `riot.im <https://riot.im/app/#/register>`_ or other Matrix clients.
+user via a Matrix client.
 
 Your new user name will be formed partly from the ``server_name``, and partly
 from a localpart you specify when you create the account. Your name will take
@@ -183,30 +179,6 @@ versions of synapse.
 
 .. _UPGRADE.rst: UPGRADE.rst
 
-
-Using PostgreSQL
-================
-
-Synapse offers two database engines:
- * `PostgreSQL <https://www.postgresql.org>`_
- * `SQLite <https://sqlite.org/>`_
-
-Almost all installations should opt to use PostgreSQL. Advantages include:
-
-* significant performance improvements due to the superior threading and
-  caching model, smarter query optimiser
-* allowing the DB to be run on separate hardware
-* allowing basic active/backup high-availability with a "hot spare" synapse
-  pointing at the same DB master, as well as enabling DB replication in
-  synapse itself.
-
-For information on how to install and use PostgreSQL, please see
-`docs/postgres.md <docs/postgres.md>`_.
-
-By default Synapse uses SQLite and in doing so trades performance for convenience.
-SQLite is only recommended in Synapse for testing purposes or for servers with
-light workloads.
-
 .. _reverse-proxy:
 
 Using a reverse proxy with Synapse
@@ -255,10 +227,9 @@ email address.
 Password reset
 ==============
 
-If a user has registered an email address to their account using an identity
-server, they can request a password-reset token via clients such as Riot.
-
-A manual password reset can be done via direct database access as follows.
+Users can reset their password through their client. Alternatively, a server admin
+can reset a users password using the `admin API <docs/admin_api/user_admin_api.rst#reset-password>`_
+or by directly editing the database as shown below.
 
 First calculate the hash of the new password::
 

changelog.d/7314.misc

@@ -0,0 +1 @@
+Allow guest access to the `GET /_matrix/client/r0/rooms/{room_id}/members` endpoint, according to MSC2689. Contributed by Awesome Technologies Innovationslabor GmbH.

changelog.d/7736.feature

@@ -0,0 +1 @@
+Add unread messages count to sync responses, as specified in [MSC2654](https://github.com/matrix-org/matrix-doc/pull/2654).

changelog.d/7899.doc

@@ -0,0 +1 @@
+Document how to set up a Client Well-Known file and fix several pieces of outdated documentation.

changelog.d/7902.feature

@@ -0,0 +1 @@
+Add option to allow server admins to join rooms which fail complexity checks. Contributed by @lugino-emeritus.

changelog.d/7936.misc

@@ -0,0 +1 @@
+Switch to the JSON implementation from the standard library and bump the minimum version of the canonicaljson library to 1.2.0.

changelog.d/7947.misc

@@ -0,0 +1 @@
+Convert various parts of the codebase to async/await.

changelog.d/7948.misc

@@ -0,0 +1 @@
+Convert various parts of the codebase to async/await.

changelog.d/7949.misc

@@ -0,0 +1 @@
+Convert various parts of the codebase to async/await.

changelog.d/7951.misc

@@ -0,0 +1 @@
+Convert various parts of the codebase to async/await.

changelog.d/7952.misc

@@ -0,0 +1 @@
+Move some database-related log lines from the default logger to the database/transaction loggers.

changelog.d/7963.misc

@@ -0,0 +1 @@
+Convert various parts of the codebase to async/await.

changelog.d/7964.feature

@@ -0,0 +1 @@
+Add an option to purge room or not with delete room admin endpoint (`POST /_synapse/admin/v1/rooms/<room_id>/delete`). Contributed by @dklimpel.

changelog.d/7965.misc

@@ -0,0 +1 @@
+Add a script to detect source code files using non-unix line terminators.

changelog.d/7970.misc

@@ -0,0 +1 @@
+Add a script to detect source code files using non-unix line terminators.

changelog.d/7971.misc

@@ -0,0 +1 @@
+Log the SAML session ID during creation.

changelog.d/7973.misc

@@ -0,0 +1 @@
+Convert various parts of the codebase to async/await.

changelog.d/7975.misc

@@ -0,0 +1 @@
+Convert various parts of the codebase to async/await.

changelog.d/7976.misc

@@ -0,0 +1 @@
+Convert various parts of the codebase to async/await.

changelog.d/7977.bugfix

@@ -0,0 +1 @@
+Fix a bug introduced in Synapse v1.7.2 which caused inaccurate membership counts in the room directory.

changelog.d/7978.bugfix

@@ -0,0 +1 @@
+Fix a long standing bug: 'Duplicate key value violates unique constraint "event_relations_id"' when message retention is configured.

changelog.d/7979.misc

@@ -0,0 +1 @@
+Switch to the JSON implementation from the standard library and bump the minimum version of the canonicaljson library to 1.2.0.

changelog.d/7980.bugfix

@@ -0,0 +1 @@
+Fix "no create event in auth events" when trying to reject invitation after inviter leaves. Bug introduced in Synapse v1.10.0.

changelog.d/7981.misc

@@ -0,0 +1 @@
+Convert various parts of the codebase to async/await.

changelog.d/7987.misc

@@ -0,0 +1 @@
+Convert various parts of the codebase to async/await.

changelog.d/7989.misc

@@ -0,0 +1 @@
+Convert various parts of the codebase to async/await.

changelog.d/7990.doc

@@ -0,0 +1 @@
+Improve workers docs.

changelog.d/7992.doc

@@ -0,0 +1 @@
+Fix typo in `docs/workers.md`.

changelog.d/7996.bugfix

@@ -0,0 +1 @@
+Fix various comments and minor discrepencies in server notices code.

changelog.d/7998.doc

@@ -0,0 +1 @@
+Add documentation for how to undo a room shutdown.

changelog.d/7999.bugfix

@@ -0,0 +1 @@
+Fix a long standing bug where HTTP HEAD requests resulted in a 400 error.

changelog.d/8001.misc

@@ -0,0 +1 @@
+Remove redundant and unreliable signature check for v1 Identity Service lookup responses.

changelog.d/8003.misc

@@ -0,0 +1 @@
+Convert various parts of the codebase to async/await.

changelog.d/8008.feature

@@ -0,0 +1 @@
+Add rate limiting to users joining rooms.

changelog.d/8025.bugfix

@@ -0,0 +1 @@
+Fix bug where state (e.g. power levels) would reset incorrectly when receiving an event from a remote server.

contrib/cmdclient/console.py

@@ -609,13 +609,15 @@ class SynapseCmd(cmd.Cmd):
 
     @defer.inlineCallbacks
     def _do_event_stream(self, timeout):
-        res = yield self.http_client.get_json(
-            self._url() + "/events",
-            {
-                "access_token": self._tok(),
-                "timeout": str(timeout),
-                "from": self.event_stream_token,
-            },
+        res = yield defer.ensureDeferred(
+            self.http_client.get_json(
+                self._url() + "/events",
+                {
+                    "access_token": self._tok(),
+                    "timeout": str(timeout),
+                    "from": self.event_stream_token,
+                },
+            )
         )
         print(json.dumps(res, indent=4))
 

debian/changelog

@@ -1,3 +1,13 @@
+matrix-synapse-py3 (1.xx.0) stable; urgency=medium
+
+  [ Synapse Packaging team ]
+  * New synapse release 1.xx.0.
+
+  [ Aaron Raimist ]
+  * Fix outdated documentation for SYNAPSE_CACHE_FACTOR
+
+ -- Synapse Packaging team <packages@matrix.org>  XXXXX
+
 matrix-synapse-py3 (1.18.0) stable; urgency=medium
 
   * New synapse release 1.18.0.

debian/matrix-synapse.default

@@ -1,2 +1,2 @@
 # Specify environment variables used when running Synapse
-# SYNAPSE_CACHE_FACTOR=1 (default)
+# SYNAPSE_CACHE_FACTOR=0.5 (default)

debian/synctl.ronn

@@ -46,19 +46,20 @@ Configuration file may be generated as follows:
 ## ENVIRONMENT
 
   * `SYNAPSE_CACHE_FACTOR`:
-    Synapse's architecture is quite RAM hungry currently - a lot of
-    recent room data and metadata is deliberately cached in RAM in
-    order to speed up common requests.  This will be improved in
-    future, but for now the easiest way to either reduce the RAM usage
-    (at the risk of slowing things down) is to set the
-    SYNAPSE_CACHE_FACTOR environment variable. Roughly speaking, a
-    SYNAPSE_CACHE_FACTOR of 1.0 will max out at around 3-4GB of
-    resident memory - this is what we currently run the matrix.org
-    on. The default setting is currently 0.1, which is probably around
-    a ~700MB footprint. You can dial it down further to 0.02 if
-    desired, which targets roughly ~512MB. Conversely you can dial it
-    up if you need performance for lots of users and have a box with a
-    lot of RAM.
+    Synapse's architecture is quite RAM hungry currently - we deliberately
+    cache a lot of recent room data and metadata in RAM in order to speed up
+    common requests. We'll improve this in the future, but for now the easiest
+    way to either reduce the RAM usage (at the risk of slowing things down)
+    is to set the almost-undocumented ``SYNAPSE_CACHE_FACTOR`` environment
+    variable. The default is 0.5, which can be decreased to reduce RAM usage
+    in memory constrained enviroments, or increased if performance starts to
+    degrade.
+
+    However, degraded performance due to a low cache factor, common on
+    machines with slow disks, often leads to explosions in memory use due
+    backlogged requests. In this case, reducing the cache factor will make
+    things worse. Instead, try increasing it drastically. 2.0 is a good
+    starting value.
 
 ## COPYRIGHT
 

docs/.sample_config_header.yaml

@@ -10,5 +10,16 @@
 # homeserver.yaml. Instead, if you are starting from scratch, please generate
 # a fresh config using Synapse by following the instructions in INSTALL.md.
 
+# Configuration options that take a time period can be set using a number
+# followed by a letter. Letters have the following meanings:
+# s = second
+# m = minute
+# h = hour
+# d = day
+# w = week
+# y = year
+# For example, setting redaction_retention_period: 5m would remove redacted
+# messages from the database after 5 minutes, rather than 5 months.
+
 ################################################################################
 

docs/admin_api/rooms.md

@@ -369,7 +369,9 @@ to the new room will have power level `-10` by default, and thus be unable to sp
 If `block` is `True` it prevents new joins to the old room.
 
 This API will remove all trace of the old room from your database after removing
-all local users.
+all local users. If `purge` is `true` (the default), all traces of the old room will
+be removed from your database after removing all local users. If you do not want
+this to happen, set `purge` to `false`.
 Depending on the amount of history being purged a call to the API may take
 several minutes or longer.
 
@@ -388,7 +390,8 @@ with a body of:
     "new_room_user_id": "@someuser:example.com",
     "room_name": "Content Violation Notification",
     "message": "Bad Room has been shutdown due to content violations on this server. Please review our Terms of Service.",
-    "block": true
+    "block": true,
+    "purge": true
 }
 ```
 
@@ -430,8 +433,10 @@ The following JSON body parameters are available:
               `new_room_user_id` in the new room. Ideally this will clearly convey why the
                original room was shut down. Defaults to `Sharing illegal content on this server
                is not permitted and rooms in violation will be blocked.`
-* `block` - Optional. If set to `true`, this room will be added to a blocking list, preventing future attempts to
-  join the room. Defaults to `false`.
+* `block` - Optional. If set to `true`, this room will be added to a blocking list, preventing
+            future attempts to join the room. Defaults to `false`.
+* `purge` - Optional. If set to `true`, it will remove all traces of the room from your database.
+            Defaults to `true`.
 
 The JSON body must not be empty. The body must be at least `{}`.
 

docs/admin_api/shutdown_room.md

@@ -33,7 +33,7 @@ You will need to authenticate with an access token for an admin user.
 * `message` - Optional. A string containing the first message that will be sent as
               `new_room_user_id` in the new room. Ideally this will clearly convey why the
                original room was shut down.
-              
+
 If not specified, the default value of `room_name` is "Content Violation
 Notification". The default value of `message` is "Sharing illegal content on
 othis server is not permitted and rooms in violation will be blocked."
@@ -72,3 +72,23 @@ Response:
     "new_room_id": "!newroomid:example.com",
 },
 ```
+
+## Undoing room shutdowns
+
+*Note*: This guide may be outdated by the time you read it. By nature of room shutdowns being performed at the database level,
+the structure can and does change without notice.
+
+First, it's important to understand that a room shutdown is very destructive. Undoing a shutdown is not as simple as pretending it
+never happened - work has to be done to move forward instead of resetting the past.
+
+1. For safety reasons, it is recommended to shut down Synapse prior to continuing.
+2. In the database, run `DELETE FROM blocked_rooms WHERE room_id = '!example:example.org';`
+   * For caution: it's recommended to run this in a transaction: `BEGIN; DELETE ...;`, verify you got 1 result, then `COMMIT;`.
+   * The room ID is the same one supplied to the shutdown room API, not the Content Violation room.
+3. Restart Synapse (required).
+
+You will have to manually handle, if you so choose, the following:
+
+* Aliases that would have been redirected to the Content Violation room.
+* Users that would have been booted from the room (and will have been force-joined to the Content Violation room).
+* Removal of the Content Violation room if desired.

docs/metrics-howto.md

@@ -27,7 +27,7 @@
     different thread to Synapse. This can make it more resilient to
     heavy load meaning metrics cannot be retrieved, and can be exposed
     to just internal networks easier. The served metrics are available
-    over HTTP only, and will be available at `/`.
+    over HTTP only, and will be available at `/_synapse/metrics`.
 
     Add a new listener to homeserver.yaml:
 

docs/postgres.md

@@ -188,6 +188,9 @@ to do step 2.
 
 It is safe to at any time kill the port script and restart it.
 
+Note that the database may take up significantly more (25% - 100% more)
+space on disk after porting to Postgres.
+
 ### Using the port script
 
 Firstly, shut down the currently running synapse server and copy its

docs/sample_config.yaml

@@ -10,6 +10,17 @@
 # homeserver.yaml. Instead, if you are starting from scratch, please generate
 # a fresh config using Synapse by following the instructions in INSTALL.md.
 
+# Configuration options that take a time period can be set using a number
+# followed by a letter. Letters have the following meanings:
+# s = second
+# m = minute
+# h = hour
+# d = day
+# w = week
+# y = year
+# For example, setting redaction_retention_period: 5m would remove redacted
+# messages from the database after 5 minutes, rather than 5 months.
+
 ################################################################################
 
 # Configuration file for Synapse.
@@ -314,6 +325,10 @@ limit_remote_rooms:
   #
   #complexity_error: "This room is too complex."
 
+  # allow server admins to join complex rooms. Default is false.
+  #
+  #admins_can_join: true
+
 # Whether to require a user to be in the room to add an alias to it.
 # Defaults to 'true'.
 #
@@ -731,6 +746,10 @@ log_config: "CONFDIR/SERVERNAME.log.config"
 #   - one for ratelimiting redactions by room admins. If this is not explicitly
 #     set then it uses the same ratelimiting as per rc_message. This is useful
 #     to allow room admins to deal with abuse quickly.
+#   - two for ratelimiting number of rooms a user can join, "local" for when
+#     users are joining rooms the server is already in (this is cheap) vs
+#     "remote" for when users are trying to join rooms not on the server (which
+#     can be more expensive)
 #
 # The defaults are as shown below.
 #
@@ -756,6 +775,14 @@ log_config: "CONFDIR/SERVERNAME.log.config"
 #rc_admin_redaction:
 #  per_second: 1
 #  burst_count: 50
+#
+#rc_joins:
+#  local:
+#    per_second: 0.1
+#    burst_count: 3
+#  remote:
+#    per_second: 0.01
+#    burst_count: 3
 
 
 # Ratelimiting settings for incoming federation
@@ -1145,24 +1172,6 @@ account_validity:
 #
 #default_identity_server: https://matrix.org
 
-# The list of identity servers trusted to verify third party
-# identifiers by this server.
-#
-# Also defines the ID server which will be called when an account is
-# deactivated (one will be picked arbitrarily).
-#
-# Note: This option is deprecated. Since v0.99.4, Synapse has tracked which identity
-# server a 3PID has been bound to. For 3PIDs bound before then, Synapse runs a
-# background migration script, informing itself that the identity server all of its
-# 3PIDs have been bound to is likely one of the below.
-#
-# As of Synapse v1.4.0, all other functionality of this option has been deprecated, and
-# it is now solely used for the purposes of the background migration script, and can be
-# removed once it has run.
-#trusted_third_party_id_servers:
-#  - matrix.org
-#  - vector.im
-
 # Handle threepid (email/phone etc) registration and password resets through a set of
 # *trusted* identity servers. Note that this allows the configured identity server to
 # reset passwords for accounts!

docs/workers.md

@@ -1,10 +1,10 @@
 # Scaling synapse via workers
 
-For small instances it recommended to run Synapse in monolith mode (the
-default). For larger instances where performance is a concern it can be helpful
-to split out functionality into multiple separate python processes. These
-processes are called 'workers', and are (eventually) intended to scale
-horizontally independently.
+For small instances it recommended to run Synapse in the default monolith mode.
+For larger instances where performance is a concern it can be helpful to split
+out functionality into multiple separate python processes. These processes are
+called 'workers', and are (eventually) intended to scale horizontally
+independently.
 
 Synapse's worker support is under active development and subject to change as
 we attempt to rapidly scale ever larger Synapse instances. However we are
@@ -23,29 +23,30 @@ The processes communicate with each other via a Synapse-specific protocol called
 feeds streams of newly written data between processes so they can be kept in
 sync with the database state.
 
-Additionally, processes may make HTTP requests to each other. Typically this is
-used for operations which need to wait for a reply - such as sending an event.
+When configured to do so, Synapse uses a 
+[Redis pub/sub channel](https://redis.io/topics/pubsub) to send the replication
+stream between all configured Synapse processes. Additionally, processes may
+make HTTP requests to each other, primarily for operations which need to wait
+for a reply ─ such as sending an event.
 
-As of Synapse v1.13.0, it is possible to configure Synapse to send replication
-via a [Redis pub/sub channel](https://redis.io/topics/pubsub), and is now the
-recommended way of configuring replication. This is an alternative to the old
-direct TCP connections to the main process: rather than all the workers
-connecting to the main process, all the workers and the main process connect to
-Redis, which relays replication commands between processes. This can give a
-significant cpu saving on the main process and will be a prerequisite for
-upcoming performance improvements.
+Redis support was added in v1.13.0 with it becoming the recommended method in
+v1.18.0. It replaced the old direct TCP connections (which is deprecated as of
+v1.18.0) to the main process. With Redis, rather than all the workers connecting
+to the main process, all the workers and the main process connect to Redis,
+which relays replication commands between processes. This can give a significant
+cpu saving on the main process and will be a prerequisite for upcoming
+performance improvements.
 
-(See the [Architectural diagram](#architectural-diagram) section at the end for
-a visualisation of what this looks like)
+See the [Architectural diagram](#architectural-diagram) section at the end for
+a visualisation of what this looks like.
 
 
 ## Setting up workers
 
 A Redis server is required to manage the communication between the processes.
-(The older direct TCP connections are now deprecated.) The Redis server
-should be installed following the normal procedure for your distribution (e.g.
-`apt install redis-server` on Debian). It is safe to use an existing Redis
-deployment if you have one.
+The Redis server should be installed following the normal procedure for your
+distribution (e.g. `apt install redis-server` on Debian). It is safe to use an
+existing Redis deployment if you have one.
 
 Once installed, check that Redis is running and accessible from the host running
 Synapse, for example by executing `echo PING | nc -q1 localhost 6379` and seeing
@@ -65,8 +66,9 @@ https://hub.docker.com/r/matrixdotorg/synapse/.
 
 To make effective use of the workers, you will need to configure an HTTP
 reverse-proxy such as nginx or haproxy, which will direct incoming requests to
-the correct worker, or to the main synapse instance. See [reverse_proxy.md](reverse_proxy.md)
-for information on setting up a reverse proxy.
+the correct worker, or to the main synapse instance. See 
+[reverse_proxy.md](reverse_proxy.md) for information on setting up a reverse
+proxy.
 
 To enable workers you should create a configuration file for each worker
 process. Each worker configuration file inherits the configuration of the shared
@@ -75,8 +77,12 @@ that worker, e.g. the HTTP listener that it provides (if any); logging
 configuration; etc.  You should minimise the number of overrides though to
 maintain a usable config.
 
-Next you need to add both a HTTP replication listener and redis config to the
-shared Synapse configuration file (`homeserver.yaml`). For example:
+
+### Shared Configuration
+
+Next you need to add both a HTTP replication listener, used for HTTP requests
+between processes, and redis config to the shared Synapse configuration file
+(`homeserver.yaml`). For example:
 
 ```yaml
 # extend the existing `listeners` section. This defines the ports that the
@@ -98,6 +104,9 @@ See the sample config for the full documentation of each option.
 Under **no circumstances** should the replication listener be exposed to the
 public internet; it has no authentication and is unencrypted.
 
+
+### Worker Configuration
+
 In the config file for each worker, you must specify the type of worker
 application (`worker_app`), and you should specify a unqiue name for the worker
 (`worker_name`). The currently available worker applications are listed below.
@@ -278,7 +287,7 @@ instance_map:
         host: localhost
         port: 8034
 
-streams_writers:
+stream_writers:
     events: event_persister1
 ```
 

scripts-dev/check_line_terminators.sh

@@ -0,0 +1,34 @@
+#!/bin/bash
+#
+# Copyright 2020 The Matrix.org Foundation C.I.C.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+#
+# This script checks that line terminators in all repository files (excluding
+# those in the .git directory) feature unix line terminators.
+#
+# Usage:
+#
+# ./check_line_terminators.sh
+#
+# The script will emit exit code 1 if any files that do not use unix line
+# terminators are found, 0 otherwise.
+
+# cd to the root of the repository
+cd `dirname $0`/..
+
+# Find and print files with non-unix line terminators
+if find . -path './.git/*' -prune -o -type f -print0 | xargs -0 grep -I -l $'\r$'; then
+    echo -e '\e[31mERROR: found files with CRLF line endings. See above.\e[39m'
+    exit 1
+fi

scripts/synapse_port_db

@@ -69,7 +69,7 @@ logger = logging.getLogger("synapse_port_db")
 
 
 BOOLEAN_COLUMNS = {
-    "events": ["processed", "outlier", "contains_url"],
+    "events": ["processed", "outlier", "contains_url", "count_as_unread"],
     "rooms": ["is_public"],
     "event_edges": ["is_state"],
     "presence_list": ["accepted"],

synapse/__init__.py

@@ -17,6 +17,7 @@
 """ This is a reference implementation of a Matrix homeserver.
 """
 
+import json
 import os
 import sys
 
@@ -25,6 +26,9 @@ if sys.version_info < (3, 5):
     print("Synapse requires Python 3.5 or above.")
     sys.exit(1)
 
+# Twisted and canonicaljson will fail to import when this file is executed to
+# get the __version__ during a fresh install. That's OK and subsequent calls to
+# actually start Synapse will import these libraries fine.
 try:
     from twisted.internet import protocol
     from twisted.internet.protocol import Factory
@@ -36,6 +40,14 @@ try:
 except ImportError:
     pass
 
+# Use the standard library json implementation instead of simplejson.
+try:
+    from canonicaljson import set_json_library
+
+    set_json_library(json)
+except ImportError:
+    pass
+
 __version__ = "1.18.0"
 
 if bool(os.environ.get("SYNAPSE_TEST_PATCH_LOG_CONTEXTS", False)):

synapse/api/auth.py

@@ -82,7 +82,7 @@ class Auth(object):
 
     @defer.inlineCallbacks
     def check_from_context(self, room_version: str, event, context, do_sig_check=True):
-        prev_state_ids = yield context.get_prev_state_ids()
+        prev_state_ids = yield defer.ensureDeferred(context.get_prev_state_ids())
         auth_events_ids = yield self.compute_auth_events(
             event, prev_state_ids, for_verification=True
         )

synapse/app/generic_worker.py

@@ -628,7 +628,7 @@ class GenericWorkerServer(HomeServer):
 
         self.get_tcp_replication().start_replication(self)
 
-    def remove_pusher(self, app_id, push_key, user_id):
+    async def remove_pusher(self, app_id, push_key, user_id):
         self.get_tcp_replication().send_remove_pusher(app_id, push_key, user_id)
 
     def build_replication_data_handler(self):

synapse/app/homeserver.py

@@ -380,13 +380,12 @@ def setup(config_options):
 
     hs.setup_master()
 
-    @defer.inlineCallbacks
-    def do_acme():
+    async def do_acme() -> bool:
         """
         Reprovision an ACME certificate, if it's required.
 
         Returns:
-            Deferred[bool]: Whether the cert has been updated.
+            Whether the cert has been updated.
         """
         acme = hs.get_acme_handler()
 
@@ -405,7 +404,7 @@ def setup(config_options):
             provision = True
 
         if provision:
-            yield acme.provision_certificate()
+            await acme.provision_certificate()
 
         return provision
 
@@ -415,7 +414,7 @@ def setup(config_options):
         Provision a certificate from ACME, if required, and reload the TLS
         certificate if it's renewed.
         """
-        reprovisioned = yield do_acme()
+        reprovisioned = yield defer.ensureDeferred(do_acme())
         if reprovisioned:
             _base.refresh_certificate(hs)
 
@@ -427,8 +426,8 @@ def setup(config_options):
                 acme = hs.get_acme_handler()
                 # Start up the webservices which we will respond to ACME
                 # challenges with, and then provision.
-                yield acme.start_listening()
-                yield do_acme()
+                yield defer.ensureDeferred(acme.start_listening())
+                yield defer.ensureDeferred(do_acme())
 
                 # Check if it needs to be reprovisioned every day.
                 hs.get_clock().looping_call(reprovision_acme, 24 * 60 * 60 * 1000)

synapse/appservice/__init__.py

@@ -15,11 +15,9 @@
 import logging
 import re
 
-from twisted.internet import defer
-
 from synapse.api.constants import EventTypes
 from synapse.types import GroupID, get_domain_from_id
-from synapse.util.caches.descriptors import cachedInlineCallbacks
+from synapse.util.caches.descriptors import cached
 
 logger = logging.getLogger(__name__)
 
@@ -43,7 +41,7 @@ class AppServiceTransaction(object):
         Args:
             as_api(ApplicationServiceApi): The API to use to send.
         Returns:
-            A Deferred which resolves to True if the transaction was sent.
+            An Awaitable which resolves to True if the transaction was sent.
         """
         return as_api.push_bulk(
             service=self.service, events=self.events, txn_id=self.id
@@ -172,8 +170,7 @@ class ApplicationService(object):
             return regex_obj["exclusive"]
         return False
 
-    @defer.inlineCallbacks
-    def _matches_user(self, event, store):
+    async def _matches_user(self, event, store):
         if not event:
             return False
 
@@ -188,12 +185,12 @@ class ApplicationService(object):
         if not store:
             return False
 
-        does_match = yield self._matches_user_in_member_list(event.room_id, store)
+        does_match = await self._matches_user_in_member_list(event.room_id, store)
         return does_match
 
-    @cachedInlineCallbacks(num_args=1, cache_context=True)
-    def _matches_user_in_member_list(self, room_id, store, cache_context):
-        member_list = yield store.get_users_in_room(
+    @cached(num_args=1, cache_context=True)
+    async def _matches_user_in_member_list(self, room_id, store, cache_context):
+        member_list = await store.get_users_in_room(
             room_id, on_invalidate=cache_context.invalidate
         )
 
@@ -208,35 +205,33 @@ class ApplicationService(object):
             return self.is_interested_in_room(event.room_id)
         return False
 
-    @defer.inlineCallbacks
-    def _matches_aliases(self, event, store):
+    async def _matches_aliases(self, event, store):
         if not store or not event:
             return False
 
-        alias_list = yield store.get_aliases_for_room(event.room_id)
+        alias_list = await store.get_aliases_for_room(event.room_id)
         for alias in alias_list:
             if self.is_interested_in_alias(alias):
                 return True
         return False
 
-    @defer.inlineCallbacks
-    def is_interested(self, event, store=None):
+    async def is_interested(self, event, store=None) -> bool:
         """Check if this service is interested in this event.
 
         Args:
             event(Event): The event to check.
             store(DataStore)
         Returns:
-            bool: True if this service would like to know about this event.
+            True if this service would like to know about this event.
         """
         # Do cheap checks first
         if self._matches_room_id(event):
             return True
 
-        if (yield self._matches_aliases(event, store)):
+        if await self._matches_aliases(event, store):
             return True
 
-        if (yield self._matches_user(event, store)):
+        if await self._matches_user(event, store):
             return True
 
         return False

synapse/appservice/api.py

@@ -93,13 +93,12 @@ class ApplicationServiceApi(SimpleHttpClient):
             hs, "as_protocol_meta", timeout_ms=HOUR_IN_MS
         )
 
-    @defer.inlineCallbacks
-    def query_user(self, service, user_id):
+    async def query_user(self, service, user_id):
         if service.url is None:
             return False
         uri = service.url + ("/users/%s" % urllib.parse.quote(user_id))
         try:
-            response = yield self.get_json(uri, {"access_token": service.hs_token})
+            response = await self.get_json(uri, {"access_token": service.hs_token})
             if response is not None:  # just an empty json object
                 return True
         except CodeMessageException as e:
@@ -110,14 +109,12 @@ class ApplicationServiceApi(SimpleHttpClient):
             logger.warning("query_user to %s threw exception %s", uri, ex)
         return False
 
-    @defer.inlineCallbacks
-    def query_alias(self, service, alias):
+    async def query_alias(self, service, alias):
         if service.url is None:
             return False
         uri = service.url + ("/rooms/%s" % urllib.parse.quote(alias))
-        response = None
         try:
-            response = yield self.get_json(uri, {"access_token": service.hs_token})
+            response = await self.get_json(uri, {"access_token": service.hs_token})
             if response is not None:  # just an empty json object
                 return True
         except CodeMessageException as e:
@@ -128,8 +125,7 @@ class ApplicationServiceApi(SimpleHttpClient):
             logger.warning("query_alias to %s threw exception %s", uri, ex)
         return False
 
-    @defer.inlineCallbacks
-    def query_3pe(self, service, kind, protocol, fields):
+    async def query_3pe(self, service, kind, protocol, fields):
         if kind == ThirdPartyEntityKind.USER:
             required_field = "userid"
         elif kind == ThirdPartyEntityKind.LOCATION:
@@ -146,7 +142,7 @@ class ApplicationServiceApi(SimpleHttpClient):
             urllib.parse.quote(protocol),
         )
         try:
-            response = yield self.get_json(uri, fields)
+            response = await self.get_json(uri, fields)
             if not isinstance(response, list):
                 logger.warning(
                     "query_3pe to %s returned an invalid response %r", uri, response
@@ -202,8 +198,7 @@ class ApplicationServiceApi(SimpleHttpClient):
         key = (service.id, protocol)
         return self.protocol_meta_cache.wrap(key, _get)
 
-    @defer.inlineCallbacks
-    def push_bulk(self, service, events, txn_id=None):
+    async def push_bulk(self, service, events, txn_id=None):
         if service.url is None:
             return True
 
@@ -218,7 +213,7 @@ class ApplicationServiceApi(SimpleHttpClient):
 
         uri = service.url + ("/transactions/%s" % urllib.parse.quote(txn_id))
         try:
-            yield self.put_json(
+            await self.put_json(
                 uri=uri,
                 json_body={"events": events},
                 args={"access_token": service.hs_token},

synapse/appservice/scheduler.py

@@ -50,8 +50,6 @@ components.
 """
 import logging
 
-from twisted.internet import defer
-
 from synapse.appservice import ApplicationServiceState
 from synapse.logging.context import run_in_background
 from synapse.metrics.background_process_metrics import run_as_background_process
@@ -73,12 +71,11 @@ class ApplicationServiceScheduler(object):
         self.txn_ctrl = _TransactionController(self.clock, self.store, self.as_api)
         self.queuer = _ServiceQueuer(self.txn_ctrl, self.clock)
 
-    @defer.inlineCallbacks
-    def start(self):
+    async def start(self):
         logger.info("Starting appservice scheduler")
 
         # check for any DOWN ASes and start recoverers for them.
-        services = yield self.store.get_appservices_by_state(
+        services = await self.store.get_appservices_by_state(
             ApplicationServiceState.DOWN
         )
 
@@ -117,8 +114,7 @@ class _ServiceQueuer(object):
             "as-sender-%s" % (service.id,), self._send_request, service
         )
 
-    @defer.inlineCallbacks
-    def _send_request(self, service):
+    async def _send_request(self, service):
         # sanity-check: we shouldn't get here if this service already has a sender
         # running.
         assert service.id not in self.requests_in_flight
@@ -130,7 +126,7 @@ class _ServiceQueuer(object):
                 if not events:
                     return
                 try:
-                    yield self.txn_ctrl.send(service, events)
+                    await self.txn_ctrl.send(service, events)
                 except Exception:
                     logger.exception("AS request failed")
         finally:
@@ -162,36 +158,33 @@ class _TransactionController(object):
         # for UTs
         self.RECOVERER_CLASS = _Recoverer
 
-    @defer.inlineCallbacks
-    def send(self, service, events):
+    async def send(self, service, events):
         try:
-            txn = yield self.store.create_appservice_txn(service=service, events=events)
-            service_is_up = yield self._is_service_up(service)
+            txn = await self.store.create_appservice_txn(service=service, events=events)
+            service_is_up = await self._is_service_up(service)
             if service_is_up:
-                sent = yield txn.send(self.as_api)
+                sent = await txn.send(self.as_api)
                 if sent:
-                    yield txn.complete(self.store)
+                    await txn.complete(self.store)
                 else:
                     run_in_background(self._on_txn_fail, service)
         except Exception:
             logger.exception("Error creating appservice transaction")
             run_in_background(self._on_txn_fail, service)
 
-    @defer.inlineCallbacks
-    def on_recovered(self, recoverer):
+    async def on_recovered(self, recoverer):
         logger.info(
             "Successfully recovered application service AS ID %s", recoverer.service.id
         )
         self.recoverers.pop(recoverer.service.id)
         logger.info("Remaining active recoverers: %s", len(self.recoverers))
-        yield self.store.set_appservice_state(
+        await self.store.set_appservice_state(
             recoverer.service, ApplicationServiceState.UP
         )
 
-    @defer.inlineCallbacks
-    def _on_txn_fail(self, service):
+    async def _on_txn_fail(self, service):
         try:
-            yield self.store.set_appservice_state(service, ApplicationServiceState.DOWN)
+            await self.store.set_appservice_state(service, ApplicationServiceState.DOWN)
             self.start_recoverer(service)
         except Exception:
             logger.exception("Error starting AS recoverer")
@@ -211,9 +204,8 @@ class _TransactionController(object):
         recoverer.recover()
         logger.info("Now %i active recoverers", len(self.recoverers))
 
-    @defer.inlineCallbacks
-    def _is_service_up(self, service):
-        state = yield self.store.get_appservice_state(service)
+    async def _is_service_up(self, service):
+        state = await self.store.get_appservice_state(service)
         return state == ApplicationServiceState.UP or state is None
 
 
@@ -254,25 +246,24 @@ class _Recoverer(object):
             self.backoff_counter += 1
         self.recover()
 
-    @defer.inlineCallbacks
-    def retry(self):
+    async def retry(self):
         logger.info("Starting retries on %s", self.service.id)
         try:
             while True:
-                txn = yield self.store.get_oldest_unsent_txn(self.service)
+                txn = await self.store.get_oldest_unsent_txn(self.service)
                 if not txn:
                     # nothing left: we're done!
-                    self.callback(self)
+                    await self.callback(self)
                     return
 
                 logger.info(
                     "Retrying transaction %s for AS ID %s", txn.id, txn.service.id
                 )
-                sent = yield txn.send(self.as_api)
+                sent = await txn.send(self.as_api)
                 if not sent:
                     break
 
-                yield txn.complete(self.store)
+                await txn.complete(self.store)
 
                 # reset the backoff counter and then process the next transaction
                 self.backoff_counter = 1

synapse/config/ratelimiting.py

@@ -93,6 +93,15 @@ class RatelimitConfig(Config):
         if rc_admin_redaction:
             self.rc_admin_redaction = RateLimitConfig(rc_admin_redaction)
 
+        self.rc_joins_local = RateLimitConfig(
+            config.get("rc_joins", {}).get("local", {}),
+            defaults={"per_second": 0.1, "burst_count": 3},
+        )
+        self.rc_joins_remote = RateLimitConfig(
+            config.get("rc_joins", {}).get("remote", {}),
+            defaults={"per_second": 0.01, "burst_count": 3},
+        )
+
     def generate_config_section(self, **kwargs):
         return """\
         ## Ratelimiting ##
@@ -118,6 +127,10 @@ class RatelimitConfig(Config):
         #   - one for ratelimiting redactions by room admins. If this is not explicitly
         #     set then it uses the same ratelimiting as per rc_message. This is useful
         #     to allow room admins to deal with abuse quickly.
+        #   - two for ratelimiting number of rooms a user can join, "local" for when
+        #     users are joining rooms the server is already in (this is cheap) vs
+        #     "remote" for when users are trying to join rooms not on the server (which
+        #     can be more expensive)
         #
         # The defaults are as shown below.
         #
@@ -143,6 +156,14 @@ class RatelimitConfig(Config):
         #rc_admin_redaction:
         #  per_second: 1
         #  burst_count: 50
+        #
+        #rc_joins:
+        #  local:
+        #    per_second: 0.1
+        #    burst_count: 3
+        #  remote:
+        #    per_second: 0.01
+        #    burst_count: 3
 
 
         # Ratelimiting settings for incoming federation

synapse/config/registration.py

@@ -333,24 +333,6 @@ class RegistrationConfig(Config):
         #
         #default_identity_server: https://matrix.org
 
-        # The list of identity servers trusted to verify third party
-        # identifiers by this server.
-        #
-        # Also defines the ID server which will be called when an account is
-        # deactivated (one will be picked arbitrarily).
-        #
-        # Note: This option is deprecated. Since v0.99.4, Synapse has tracked which identity
-        # server a 3PID has been bound to. For 3PIDs bound before then, Synapse runs a
-        # background migration script, informing itself that the identity server all of its
-        # 3PIDs have been bound to is likely one of the below.
-        #
-        # As of Synapse v1.4.0, all other functionality of this option has been deprecated, and
-        # it is now solely used for the purposes of the background migration script, and can be
-        # removed once it has run.
-        #trusted_third_party_id_servers:
-        #  - matrix.org
-        #  - vector.im
-
         # Handle threepid (email/phone etc) registration and password resets through a set of
         # *trusted* identity servers. Note that this allows the configured identity server to
         # reset passwords for accounts!

synapse/config/server.py

@@ -439,6 +439,9 @@ class ServerConfig(Config):
                 validator=attr.validators.instance_of(str),
                 default=ROOM_COMPLEXITY_TOO_GREAT,
             )
+            admins_can_join = attr.ib(
+                validator=attr.validators.instance_of(bool), default=False
+            )
 
         self.limit_remote_rooms = LimitRemoteRoomsConfig(
             **(config.get("limit_remote_rooms") or {})
@@ -893,6 +896,10 @@ class ServerConfig(Config):
           #
           #complexity_error: "This room is too complex."
 
+          # allow server admins to join complex rooms. Default is false.
+          #
+          #admins_can_join: true
+
         # Whether to require a user to be in the room to add an alias to it.
         # Defaults to 'true'.
         #

synapse/crypto/keyring.py

@@ -223,8 +223,7 @@ class Keyring(object):
 
         return results
 
-    @defer.inlineCallbacks
-    def _start_key_lookups(self, verify_requests):
+    async def _start_key_lookups(self, verify_requests):
         """Sets off the key fetches for each verify request
 
         Once each fetch completes, verify_request.key_ready will be resolved.
@@ -245,7 +244,7 @@ class Keyring(object):
                 server_to_request_ids.setdefault(server_name, set()).add(request_id)
 
             # Wait for any previous lookups to complete before proceeding.
-            yield self.wait_for_previous_lookups(server_to_request_ids.keys())
+            await self.wait_for_previous_lookups(server_to_request_ids.keys())
 
             # take out a lock on each of the servers by sticking a Deferred in
             # key_downloads
@@ -283,15 +282,14 @@ class Keyring(object):
         except Exception:
             logger.exception("Error starting key lookups")
 
-    @defer.inlineCallbacks
-    def wait_for_previous_lookups(self, server_names):
+    async def wait_for_previous_lookups(self, server_names) -> None:
         """Waits for any previous key lookups for the given servers to finish.
 
         Args:
             server_names (Iterable[str]): list of servers which we want to look up
 
         Returns:
-            Deferred[None]: resolves once all key lookups for the given servers have
+            Resolves once all key lookups for the given servers have
                 completed. Follows the synapse rules of logcontext preservation.
         """
         loop_count = 1
@@ -309,7 +307,7 @@ class Keyring(object):
                 loop_count,
             )
             with PreserveLoggingContext():
-                yield defer.DeferredList((w[1] for w in wait_on))
+                await defer.DeferredList((w[1] for w in wait_on))
 
             loop_count += 1
 
@@ -326,44 +324,44 @@ class Keyring(object):
 
         remaining_requests = {rq for rq in verify_requests if not rq.key_ready.called}
 
-        @defer.inlineCallbacks
-        def do_iterations():
-            with Measure(self.clock, "get_server_verify_keys"):
-                for f in self._key_fetchers:
-                    if not remaining_requests:
-                        return
-                    yield self._attempt_key_fetches_with_fetcher(f, remaining_requests)
-
-                # look for any requests which weren't satisfied
-                with PreserveLoggingContext():
-                    for verify_request in remaining_requests:
-                        verify_request.key_ready.errback(
-                            SynapseError(
-                                401,
-                                "No key for %s with ids in %s (min_validity %i)"
-                                % (
-                                    verify_request.server_name,
-                                    verify_request.key_ids,
-                                    verify_request.minimum_valid_until_ts,
-                                ),
-                                Codes.UNAUTHORIZED,
-                            )
+        async def do_iterations():
+            try:
+                with Measure(self.clock, "get_server_verify_keys"):
+                    for f in self._key_fetchers:
+                        if not remaining_requests:
+                            return
+                        await self._attempt_key_fetches_with_fetcher(
+                            f, remaining_requests
                         )
 
-        def on_err(err):
-            # we don't really expect to get here, because any errors should already
-            # have been caught and logged. But if we do, let's log the error and make
-            # sure that all of the deferreds are resolved.
-            logger.error("Unexpected error in _get_server_verify_keys: %s", err)
-            with PreserveLoggingContext():
-                for verify_request in remaining_requests:
-                    if not verify_request.key_ready.called:
-                        verify_request.key_ready.errback(err)
+                    # look for any requests which weren't satisfied
+                    with PreserveLoggingContext():
+                        for verify_request in remaining_requests:
+                            verify_request.key_ready.errback(
+                                SynapseError(
+                                    401,
+                                    "No key for %s with ids in %s (min_validity %i)"
+                                    % (
+                                        verify_request.server_name,
+                                        verify_request.key_ids,
+                                        verify_request.minimum_valid_until_ts,
+                                    ),
+                                    Codes.UNAUTHORIZED,
+                                )
+                            )
+            except Exception as err:
+                # we don't really expect to get here, because any errors should already
+                # have been caught and logged. But if we do, let's log the error and make
+                # sure that all of the deferreds are resolved.
+                logger.error("Unexpected error in _get_server_verify_keys: %s", err)
+                with PreserveLoggingContext():
+                    for verify_request in remaining_requests:
+                        if not verify_request.key_ready.called:
+                            verify_request.key_ready.errback(err)
 
-        run_in_background(do_iterations).addErrback(on_err)
+        run_in_background(do_iterations)
 
-    @defer.inlineCallbacks
-    def _attempt_key_fetches_with_fetcher(self, fetcher, remaining_requests):
+    async def _attempt_key_fetches_with_fetcher(self, fetcher, remaining_requests):
         """Use a key fetcher to attempt to satisfy some key requests
 
         Args:
@@ -390,7 +388,7 @@ class Keyring(object):
                     verify_request.minimum_valid_until_ts,
                 )
 
-        results = yield fetcher.get_keys(missing_keys)
+        results = await fetcher.get_keys(missing_keys)
 
         completed = []
         for verify_request in remaining_requests:
@@ -423,7 +421,7 @@ class Keyring(object):
 
 
 class KeyFetcher(object):
-    def get_keys(self, keys_to_fetch):
+    async def get_keys(self, keys_to_fetch):
         """
         Args:
             keys_to_fetch (dict[str, dict[str, int]]):
@@ -442,8 +440,7 @@ class StoreKeyFetcher(KeyFetcher):
     def __init__(self, hs):
         self.store = hs.get_datastore()
 
-    @defer.inlineCallbacks
-    def get_keys(self, keys_to_fetch):
+    async def get_keys(self, keys_to_fetch):
         """see KeyFetcher.get_keys"""
 
         keys_to_fetch = (
@@ -452,7 +449,7 @@ class StoreKeyFetcher(KeyFetcher):
             for key_id in keys_for_server.keys()
         )
 
-        res = yield self.store.get_server_verify_keys(keys_to_fetch)
+        res = await self.store.get_server_verify_keys(keys_to_fetch)
         keys = {}
         for (server_name, key_id), key in res.items():
             keys.setdefault(server_name, {})[key_id] = key
@@ -464,8 +461,7 @@ class BaseV2KeyFetcher(object):
         self.store = hs.get_datastore()
         self.config = hs.get_config()
 
-    @defer.inlineCallbacks
-    def process_v2_response(self, from_server, response_json, time_added_ms):
+    async def process_v2_response(self, from_server, response_json, time_added_ms):
         """Parse a 'Server Keys' structure from the result of a /key request
 
         This is used to parse either the entirety of the response from
@@ -537,7 +533,7 @@ class BaseV2KeyFetcher(object):
 
         key_json_bytes = encode_canonical_json(response_json)
 
-        yield make_deferred_yieldable(
+        await make_deferred_yieldable(
             defer.gatherResults(
                 [
                     run_in_background(
@@ -567,14 +563,12 @@ class PerspectivesKeyFetcher(BaseV2KeyFetcher):
         self.client = hs.get_http_client()
         self.key_servers = self.config.key_servers
 
-    @defer.inlineCallbacks
-    def get_keys(self, keys_to_fetch):
+    async def get_keys(self, keys_to_fetch):
         """see KeyFetcher.get_keys"""
 
-        @defer.inlineCallbacks
-        def get_key(key_server):
+        async def get_key(key_server):
             try:
-                result = yield self.get_server_verify_key_v2_indirect(
+                result = await self.get_server_verify_key_v2_indirect(
                     keys_to_fetch, key_server
                 )
                 return result
@@ -592,7 +586,7 @@ class PerspectivesKeyFetcher(BaseV2KeyFetcher):
 
             return {}
 
-        results = yield make_deferred_yieldable(
+        results = await make_deferred_yieldable(
             defer.gatherResults(
                 [run_in_background(get_key, server) for server in self.key_servers],
                 consumeErrors=True,
@@ -606,8 +600,7 @@ class PerspectivesKeyFetcher(BaseV2KeyFetcher):
 
         return union_of_keys
 
-    @defer.inlineCallbacks
-    def get_server_verify_key_v2_indirect(self, keys_to_fetch, key_server):
+    async def get_server_verify_key_v2_indirect(self, keys_to_fetch, key_server):
         """
         Args:
             keys_to_fetch (dict[str, dict[str, int]]):
@@ -617,7 +610,7 @@ class PerspectivesKeyFetcher(BaseV2KeyFetcher):
                 the keys
 
         Returns:
-            Deferred[dict[str, dict[str, synapse.storage.keys.FetchKeyResult]]]: map
+            dict[str, dict[str, synapse.storage.keys.FetchKeyResult]]: map
                 from server_name -> key_id -> FetchKeyResult
 
         Raises:
@@ -632,7 +625,7 @@ class PerspectivesKeyFetcher(BaseV2KeyFetcher):
         )
 
         try:
-            query_response = yield self.client.post_json(
+            query_response = await self.client.post_json(
                 destination=perspective_name,
                 path="/_matrix/key/v2/query",
                 data={
@@ -668,7 +661,7 @@ class PerspectivesKeyFetcher(BaseV2KeyFetcher):
             try:
                 self._validate_perspectives_response(key_server, response)
 
-                processed_response = yield self.process_v2_response(
+                processed_response = await self.process_v2_response(
                     perspective_name, response, time_added_ms=time_now_ms
                 )
             except KeyLookupError as e:
@@ -687,7 +680,7 @@ class PerspectivesKeyFetcher(BaseV2KeyFetcher):
             )
             keys.setdefault(server_name, {}).update(processed_response)
 
-        yield self.store.store_server_verify_keys(
+        await self.store.store_server_verify_keys(
             perspective_name, time_now_ms, added_keys
         )
 
@@ -739,24 +732,23 @@ class ServerKeyFetcher(BaseV2KeyFetcher):
         self.clock = hs.get_clock()
         self.client = hs.get_http_client()
 
-    def get_keys(self, keys_to_fetch):
+    async def get_keys(self, keys_to_fetch):
         """
         Args:
             keys_to_fetch (dict[str, iterable[str]]):
                 the keys to be fetched. server_name -> key_ids
 
         Returns:
-            Deferred[dict[str, dict[str, synapse.storage.keys.FetchKeyResult|None]]]:
+            dict[str, dict[str, synapse.storage.keys.FetchKeyResult|None]]:
                 map from server_name -> key_id -> FetchKeyResult
         """
 
         results = {}
 
-        @defer.inlineCallbacks
-        def get_key(key_to_fetch_item):
+        async def get_key(key_to_fetch_item):
             server_name, key_ids = key_to_fetch_item
             try:
-                keys = yield self.get_server_verify_key_v2_direct(server_name, key_ids)
+                keys = await self.get_server_verify_key_v2_direct(server_name, key_ids)
                 results[server_name] = keys
             except KeyLookupError as e:
                 logger.warning(
@@ -765,12 +757,11 @@ class ServerKeyFetcher(BaseV2KeyFetcher):
             except Exception:
                 logger.exception("Error getting keys %s from %s", key_ids, server_name)
 
-        return yieldable_gather_results(get_key, keys_to_fetch.items()).addCallback(
-            lambda _: results
-        )
+        return await yieldable_gather_results(
+            get_key, keys_to_fetch.items()
+        ).addCallback(lambda _: results)
 
-    @defer.inlineCallbacks
-    def get_server_verify_key_v2_direct(self, server_name, key_ids):
+    async def get_server_verify_key_v2_direct(self, server_name, key_ids):
         """
 
         Args:
@@ -792,7 +783,7 @@ class ServerKeyFetcher(BaseV2KeyFetcher):
 
             time_now_ms = self.clock.time_msec()
             try:
-                response = yield self.client.get_json(
+                response = await self.client.get_json(
                     destination=server_name,
                     path="/_matrix/key/v2/server/"
                     + urllib.parse.quote(requested_key_id),
@@ -823,12 +814,12 @@ class ServerKeyFetcher(BaseV2KeyFetcher):
                     % (server_name, response["server_name"])
                 )
 
-            response_keys = yield self.process_v2_response(
+            response_keys = await self.process_v2_response(
                 from_server=server_name,
                 response_json=response,
                 time_added_ms=time_now_ms,
             )
-            yield self.store.store_server_verify_keys(
+            await self.store.store_server_verify_keys(
                 server_name,
                 time_now_ms,
                 ((server_name, key_id, key) for key_id, key in response_keys.items()),
@@ -838,22 +829,18 @@ class ServerKeyFetcher(BaseV2KeyFetcher):
         return keys
 
 
-@defer.inlineCallbacks
-def _handle_key_deferred(verify_request):
+async def _handle_key_deferred(verify_request) -> None:
     """Waits for the key to become available, and then performs a verification
 
     Args:
         verify_request (VerifyJsonRequest):
 
-    Returns:
-        Deferred[None]
-
     Raises:
         SynapseError if there was a problem performing the verification
     """
     server_name = verify_request.server_name
     with PreserveLoggingContext():
-        _, key_id, verify_key = yield verify_request.key_ready
+        _, key_id, verify_key = await verify_request.key_ready
 
     json_object = verify_request.json_object
 

synapse/events/builder.py

@@ -17,8 +17,6 @@ from typing import Optional
 import attr
 from nacl.signing import SigningKey
 
-from twisted.internet import defer
-
 from synapse.api.constants import MAX_DEPTH
 from synapse.api.errors import UnsupportedRoomVersionError
 from synapse.api.room_versions import (
@@ -95,31 +93,30 @@ class EventBuilder(object):
     def is_state(self):
         return self._state_key is not None
 
-    @defer.inlineCallbacks
-    def build(self, prev_event_ids):
+    async def build(self, prev_event_ids):
         """Transform into a fully signed and hashed event
 
         Args:
             prev_event_ids (list[str]): The event IDs to use as the prev events
 
         Returns:
-            Deferred[FrozenEvent]
+            FrozenEvent
         """
 
-        state_ids = yield defer.ensureDeferred(
-            self._state.get_current_state_ids(self.room_id, prev_event_ids)
+        state_ids = await self._state.get_current_state_ids(
+            self.room_id, prev_event_ids
         )
-        auth_ids = yield self._auth.compute_auth_events(self, state_ids)
+        auth_ids = await self._auth.compute_auth_events(self, state_ids)
 
         format_version = self.room_version.event_format
         if format_version == EventFormatVersions.V1:
-            auth_events = yield self._store.add_event_hashes(auth_ids)
-            prev_events = yield self._store.add_event_hashes(prev_event_ids)
+            auth_events = await self._store.add_event_hashes(auth_ids)
+            prev_events = await self._store.add_event_hashes(prev_event_ids)
         else:
             auth_events = auth_ids
             prev_events = prev_event_ids
 
-        old_depth = yield self._store.get_max_depth_of(prev_event_ids)
+        old_depth = await self._store.get_max_depth_of(prev_event_ids)
         depth = old_depth + 1
 
         # we cap depth of generated events, to ensure that they are not

synapse/events/snapshot.py

@@ -12,17 +12,19 @@
 # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 # See the License for the specific language governing permissions and
 # limitations under the License.
-from typing import Optional, Union
+from typing import TYPE_CHECKING, Optional, Union
 
 import attr
 from frozendict import frozendict
 
-from twisted.internet import defer
-
 from synapse.appservice import ApplicationService
+from synapse.events import EventBase
 from synapse.logging.context import make_deferred_yieldable, run_in_background
 from synapse.types import StateMap
 
+if TYPE_CHECKING:
+    from synapse.storage.data_stores.main import DataStore
+
 
 @attr.s(slots=True)
 class EventContext:
@@ -129,8 +131,7 @@ class EventContext:
             delta_ids=delta_ids,
         )
 
-    @defer.inlineCallbacks
-    def serialize(self, event, store):
+    async def serialize(self, event: EventBase, store: "DataStore") -> dict:
         """Converts self to a type that can be serialized as JSON, and then
         deserialized by `deserialize`
 
@@ -146,7 +147,7 @@ class EventContext:
         # the prev_state_ids, so if we're a state event we include the event
         # id that we replaced in the state.
         if event.is_state():
-            prev_state_ids = yield self.get_prev_state_ids()
+            prev_state_ids = await self.get_prev_state_ids()
             prev_state_id = prev_state_ids.get((event.type, event.state_key))
         else:
             prev_state_id = None
@@ -214,8 +215,7 @@ class EventContext:
 
         return self._state_group
 
-    @defer.inlineCallbacks
-    def get_current_state_ids(self):
+    async def get_current_state_ids(self) -> Optional[StateMap[str]]:
         """
         Gets the room state map, including this event - ie, the state in ``state_group``
 
@@ -224,32 +224,31 @@ class EventContext:
         ``rejected`` is set.
 
         Returns:
-            Deferred[dict[(str, str), str]|None]: Returns None if state_group
-                is None, which happens when the associated event is an outlier.
+            Returns None if state_group is None, which happens when the associated
+            event is an outlier.
 
-                Maps a (type, state_key) to the event ID of the state event matching
-                this tuple.
+            Maps a (type, state_key) to the event ID of the state event matching
+            this tuple.
         """
         if self.rejected:
             raise RuntimeError("Attempt to access state_ids of rejected event")
 
-        yield self._ensure_fetched()
+        await self._ensure_fetched()
         return self._current_state_ids
 
-    @defer.inlineCallbacks
-    def get_prev_state_ids(self):
+    async def get_prev_state_ids(self):
         """
         Gets the room state map, excluding this event.
 
         For a non-state event, this will be the same as get_current_state_ids().
 
         Returns:
-            Deferred[dict[(str, str), str]|None]: Returns None if state_group
+            dict[(str, str), str]|None: Returns None if state_group
                 is None, which happens when the associated event is an outlier.
                 Maps a (type, state_key) to the event ID of the state event matching
                 this tuple.
         """
-        yield self._ensure_fetched()
+        await self._ensure_fetched()
         return self._prev_state_ids
 
     def get_cached_current_state_ids(self):
@@ -269,8 +268,8 @@ class EventContext:
 
         return self._current_state_ids
 
-    def _ensure_fetched(self):
-        return defer.succeed(None)
+    async def _ensure_fetched(self):
+        return None
 
 
 @attr.s(slots=True)
@@ -303,21 +302,20 @@ class _AsyncEventContextImpl(EventContext):
     _event_state_key = attr.ib(default=None)
     _fetching_state_deferred = attr.ib(default=None)
 
-    def _ensure_fetched(self):
+    async def _ensure_fetched(self):
         if not self._fetching_state_deferred:
             self._fetching_state_deferred = run_in_background(self._fill_out_state)
 
-        return make_deferred_yieldable(self._fetching_state_deferred)
+        return await make_deferred_yieldable(self._fetching_state_deferred)
 
-    @defer.inlineCallbacks
-    def _fill_out_state(self):
+    async def _fill_out_state(self):
         """Called to populate the _current_state_ids and _prev_state_ids
         attributes by loading from the database.
         """
         if self.state_group is None:
             return
 
-        self._current_state_ids = yield self._storage.state.get_state_ids_for_group(
+        self._current_state_ids = await self._storage.state.get_state_ids_for_group(
             self.state_group
         )
         if self._event_state_key is not None:

synapse/events/third_party_rules.py

@@ -13,7 +13,9 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-from twisted.internet import defer
+from synapse.events import EventBase
+from synapse.events.snapshot import EventContext
+from synapse.types import Requester
 
 
 class ThirdPartyEventRules(object):
@@ -39,76 +41,79 @@ class ThirdPartyEventRules(object):
                 config=config, http_client=hs.get_simple_http_client()
             )
 
-    @defer.inlineCallbacks
-    def check_event_allowed(self, event, context):
+    async def check_event_allowed(
+        self, event: EventBase, context: EventContext
+    ) -> bool:
         """Check if a provided event should be allowed in the given context.
 
         Args:
-            event (synapse.events.EventBase): The event to be checked.
-            context (synapse.events.snapshot.EventContext): The context of the event.
+            event: The event to be checked.
+            context: The context of the event.
 
         Returns:
-            defer.Deferred[bool]: True if the event should be allowed, False if not.
+            True if the event should be allowed, False if not.
         """
         if self.third_party_rules is None:
             return True
 
-        prev_state_ids = yield context.get_prev_state_ids()
+        prev_state_ids = await context.get_prev_state_ids()
 
         # Retrieve the state events from the database.
         state_events = {}
         for key, event_id in prev_state_ids.items():
-            state_events[key] = yield self.store.get_event(event_id, allow_none=True)
+            state_events[key] = await self.store.get_event(event_id, allow_none=True)
 
-        ret = yield self.third_party_rules.check_event_allowed(event, state_events)
+        ret = await self.third_party_rules.check_event_allowed(event, state_events)
         return ret
 
-    @defer.inlineCallbacks
-    def on_create_room(self, requester, config, is_requester_admin):
+    async def on_create_room(
+        self, requester: Requester, config: dict, is_requester_admin: bool
+    ) -> bool:
         """Intercept requests to create room to allow, deny or update the
         request config.
 
         Args:
-            requester (Requester)
-            config (dict): The creation config from the client.
-            is_requester_admin (bool): If the requester is an admin
+            requester
+            config: The creation config from the client.
+            is_requester_admin: If the requester is an admin
 
         Returns:
-            defer.Deferred[bool]: Whether room creation is allowed or denied.
+            Whether room creation is allowed or denied.
         """
 
         if self.third_party_rules is None:
             return True
 
-        ret = yield self.third_party_rules.on_create_room(
+        ret = await self.third_party_rules.on_create_room(
             requester, config, is_requester_admin
         )
         return ret
 
-    @defer.inlineCallbacks
-    def check_threepid_can_be_invited(self, medium, address, room_id):
+    async def check_threepid_can_be_invited(
+        self, medium: str, address: str, room_id: str
+    ) -> bool:
         """Check if a provided 3PID can be invited in the given room.
 
         Args:
-            medium (str): The 3PID's medium.
-            address (str): The 3PID's address.
-            room_id (str): The room we want to invite the threepid to.
+            medium: The 3PID's medium.
+            address: The 3PID's address.
+            room_id: The room we want to invite the threepid to.
 
         Returns:
-            defer.Deferred[bool], True if the 3PID can be invited, False if not.
+            True if the 3PID can be invited, False if not.
         """
 
         if self.third_party_rules is None:
             return True
 
-        state_ids = yield self.store.get_filtered_current_state_ids(room_id)
-        room_state_events = yield self.store.get_events(state_ids.values())
+        state_ids = await self.store.get_filtered_current_state_ids(room_id)
+        room_state_events = await self.store.get_events(state_ids.values())
 
         state_events = {}
         for key, event_id in state_ids.items():
             state_events[key] = room_state_events[event_id]
 
-        ret = yield self.third_party_rules.check_threepid_can_be_invited(
+        ret = await self.third_party_rules.check_threepid_can_be_invited(
             medium, address, state_events
         )
         return ret

synapse/events/utils.py

@@ -18,8 +18,6 @@ from typing import Any, Mapping, Union
 
 from frozendict import frozendict
 
-from twisted.internet import defer
-
 from synapse.api.constants import EventTypes, RelationTypes
 from synapse.api.errors import Codes, SynapseError
 from synapse.api.room_versions import RoomVersion
@@ -337,8 +335,9 @@ class EventClientSerializer(object):
             hs.config.experimental_msc1849_support_enabled
         )
 
-    @defer.inlineCallbacks
-    def serialize_event(self, event, time_now, bundle_aggregations=True, **kwargs):
+    async def serialize_event(
+        self, event, time_now, bundle_aggregations=True, **kwargs
+    ):
         """Serializes a single event.
 
         Args:
@@ -348,7 +347,7 @@ class EventClientSerializer(object):
             **kwargs: Arguments to pass to `serialize_event`
 
         Returns:
-            Deferred[dict]: The serialized event
+            dict: The serialized event
         """
         # To handle the case of presence events and the like
         if not isinstance(event, EventBase):
@@ -363,8 +362,8 @@ class EventClientSerializer(object):
         if not event.internal_metadata.is_redacted() and (
             self.experimental_msc1849_support_enabled and bundle_aggregations
         ):
-            annotations = yield self.store.get_aggregation_groups_for_event(event_id)
-            references = yield self.store.get_relations_for_event(
+            annotations = await self.store.get_aggregation_groups_for_event(event_id)
+            references = await self.store.get_relations_for_event(
                 event_id, RelationTypes.REFERENCE, direction="f"
             )
 
@@ -378,7 +377,7 @@ class EventClientSerializer(object):
 
             edit = None
             if event.type == EventTypes.Message:
-                edit = yield self.store.get_applicable_edit(event_id)
+                edit = await self.store.get_applicable_edit(event_id)
 
             if edit:
                 # If there is an edit replace the content, preserving existing

synapse/federation/federation_client.py

@@ -135,7 +135,7 @@ class FederationClient(FederationBase):
                 and try the request anyway.
 
         Returns:
-            a Deferred which will eventually yield a JSON object from the
+            a Awaitable which will eventually yield a JSON object from the
             response
         """
         sent_queries_counter.labels(query_type).inc()
@@ -157,7 +157,7 @@ class FederationClient(FederationBase):
             content (dict): The query content.
 
         Returns:
-            a Deferred which will eventually yield a JSON object from the
+            an Awaitable which will eventually yield a JSON object from the
             response
         """
         sent_queries_counter.labels("client_device_keys").inc()
@@ -180,7 +180,7 @@ class FederationClient(FederationBase):
             content (dict): The query content.
 
         Returns:
-            a Deferred which will eventually yield a JSON object from the
+            an Awaitable which will eventually yield a JSON object from the
             response
         """
         sent_queries_counter.labels("client_one_time_keys").inc()
@@ -900,7 +900,7 @@ class FederationClient(FederationBase):
                 party instance
 
         Returns:
-            Deferred[Dict[str, Any]]: The response from the remote server, or None if
+            Awaitable[Dict[str, Any]]: The response from the remote server, or None if
             `remote_server` is the same as the local server_name
 
         Raises:

synapse/federation/sender/__init__.py

@@ -288,8 +288,7 @@ class FederationSender(object):
         for destination in destinations:
             self._get_per_destination_queue(destination).send_pdu(pdu, order)
 
-    @defer.inlineCallbacks
-    def send_read_receipt(self, receipt: ReadReceipt):
+    async def send_read_receipt(self, receipt: ReadReceipt) -> None:
         """Send a RR to any other servers in the room
 
         Args:
@@ -330,9 +329,7 @@ class FederationSender(object):
         room_id = receipt.room_id
 
         # Work out which remote servers should be poked and poke them.
-        domains = yield defer.ensureDeferred(
-            self.state.get_current_hosts_in_room(room_id)
-        )
+        domains = await self.state.get_current_hosts_in_room(room_id)
         domains = [
             d
             for d in domains
@@ -387,8 +384,7 @@ class FederationSender(object):
             queue.flush_read_receipts_for_room(room_id)
 
     @preserve_fn  # the caller should not yield on this
-    @defer.inlineCallbacks
-    def send_presence(self, states: List[UserPresenceState]):
+    async def send_presence(self, states: List[UserPresenceState]):
         """Send the new presence states to the appropriate destinations.
 
         This actually queues up the presence states ready for sending and
@@ -423,7 +419,7 @@ class FederationSender(object):
                 if not states_map:
                     break
 
-                yield self._process_presence_inner(list(states_map.values()))
+                await self._process_presence_inner(list(states_map.values()))
         except Exception:
             logger.exception("Error sending presence states to servers")
         finally:
@@ -450,14 +446,11 @@ class FederationSender(object):
             self._get_per_destination_queue(destination).send_presence(states)
 
     @measure_func("txnqueue._process_presence")
-    @defer.inlineCallbacks
-    def _process_presence_inner(self, states: List[UserPresenceState]):
+    async def _process_presence_inner(self, states: List[UserPresenceState]):
         """Given a list of states populate self.pending_presence_by_dest and
         poke to send a new transaction to each destination
         """
-        hosts_and_states = yield defer.ensureDeferred(
-            get_interested_remotes(self.store, states, self.state)
-        )
+        hosts_and_states = await get_interested_remotes(self.store, states, self.state)
 
         for destinations, states in hosts_and_states:
             for destination in destinations:

synapse/federation/transport/client.py

@@ -18,8 +18,6 @@ import logging
 import urllib
 from typing import Any, Dict, Optional
 
-from twisted.internet import defer
-
 from synapse.api.constants import Membership
 from synapse.api.errors import Codes, HttpResponseException, SynapseError
 from synapse.api.urls import (
@@ -51,7 +49,7 @@ class TransportLayerClient(object):
             event_id (str): The event we want the context at.
 
         Returns:
-            Deferred: Results in a dict received from the remote homeserver.
+            Awaitable: Results in a dict received from the remote homeserver.
         """
         logger.debug("get_room_state_ids dest=%s, room=%s", destination, room_id)
 
@@ -75,7 +73,7 @@ class TransportLayerClient(object):
                 giving up. None indicates no timeout.
 
         Returns:
-            Deferred: Results in a dict received from the remote homeserver.
+            Awaitable: Results in a dict received from the remote homeserver.
         """
         logger.debug("get_pdu dest=%s, event_id=%s", destination, event_id)
 
@@ -96,7 +94,7 @@ class TransportLayerClient(object):
             limit (int)
 
         Returns:
-            Deferred: Results in a dict received from the remote homeserver.
+            Awaitable: Results in a dict received from the remote homeserver.
         """
         logger.debug(
             "backfill dest=%s, room_id=%s, event_tuples=%r, limit=%s",
@@ -118,16 +116,15 @@ class TransportLayerClient(object):
             destination, path=path, args=args, try_trailing_slash_on_400=True
         )
 
-    @defer.inlineCallbacks
     @log_function
-    def send_transaction(self, transaction, json_data_callback=None):
+    async def send_transaction(self, transaction, json_data_callback=None):
         """ Sends the given Transaction to its destination
 
         Args:
             transaction (Transaction)
 
         Returns:
-            Deferred: Succeeds when we get a 2xx HTTP response. The result
+            Succeeds when we get a 2xx HTTP response. The result
             will be the decoded JSON body.
 
             Fails with ``HTTPRequestException`` if we get an HTTP response
@@ -154,7 +151,7 @@ class TransportLayerClient(object):
 
         path = _create_v1_path("/send/%s", transaction.transaction_id)
 
-        response = yield self.client.put_json(
+        response = await self.client.put_json(
             transaction.destination,
             path=path,
             data=json_data,
@@ -166,14 +163,13 @@ class TransportLayerClient(object):
 
         return response
 
-    @defer.inlineCallbacks
     @log_function
-    def make_query(
+    async def make_query(
         self, destination, query_type, args, retry_on_dns_fail, ignore_backoff=False
     ):
         path = _create_v1_path("/query/%s", query_type)
 
-        content = yield self.client.get_json(
+        content = await self.client.get_json(
             destination=destination,
             path=path,
             args=args,
@@ -184,9 +180,10 @@ class TransportLayerClient(object):
 
         return content
 
-    @defer.inlineCallbacks
     @log_function
-    def make_membership_event(self, destination, room_id, user_id, membership, params):
+    async def make_membership_event(
+        self, destination, room_id, user_id, membership, params
+    ):
         """Asks a remote server to build and sign us a membership event
 
         Note that this does not append any events to any graphs.
@@ -200,7 +197,7 @@ class TransportLayerClient(object):
                 request.
 
         Returns:
-            Deferred: Succeeds when we get a 2xx HTTP response. The result
+            Succeeds when we get a 2xx HTTP response. The result
             will be the decoded JSON body (ie, the new event).
 
             Fails with ``HTTPRequestException`` if we get an HTTP response
@@ -231,7 +228,7 @@ class TransportLayerClient(object):
             ignore_backoff = True
             retry_on_dns_fail = True
 
-        content = yield self.client.get_json(
+        content = await self.client.get_json(
             destination=destination,
             path=path,
             args=params,
@@ -242,34 +239,31 @@ class TransportLayerClient(object):
 
         return content
 
-    @defer.inlineCallbacks
     @log_function
-    def send_join_v1(self, destination, room_id, event_id, content):
+    async def send_join_v1(self, destination, room_id, event_id, content):
         path = _create_v1_path("/send_join/%s/%s", room_id, event_id)
 
-        response = yield self.client.put_json(
+        response = await self.client.put_json(
             destination=destination, path=path, data=content
         )
 
         return response
 
-    @defer.inlineCallbacks
     @log_function
-    def send_join_v2(self, destination, room_id, event_id, content):
+    async def send_join_v2(self, destination, room_id, event_id, content):
         path = _create_v2_path("/send_join/%s/%s", room_id, event_id)
 
-        response = yield self.client.put_json(
+        response = await self.client.put_json(
             destination=destination, path=path, data=content
         )
 
         return response
 
-    @defer.inlineCallbacks
     @log_function
-    def send_leave_v1(self, destination, room_id, event_id, content):
+    async def send_leave_v1(self, destination, room_id, event_id, content):
         path = _create_v1_path("/send_leave/%s/%s", room_id, event_id)
 
-        response = yield self.client.put_json(
+        response = await self.client.put_json(
             destination=destination,
             path=path,
             data=content,
@@ -282,12 +276,11 @@ class TransportLayerClient(object):
 
         return response
 
-    @defer.inlineCallbacks
     @log_function
-    def send_leave_v2(self, destination, room_id, event_id, content):
+    async def send_leave_v2(self, destination, room_id, event_id, content):
         path = _create_v2_path("/send_leave/%s/%s", room_id, event_id)
 
-        response = yield self.client.put_json(
+        response = await self.client.put_json(
             destination=destination,
             path=path,
             data=content,
@@ -300,31 +293,28 @@ class TransportLayerClient(object):
 
         return response
 
-    @defer.inlineCallbacks
     @log_function
-    def send_invite_v1(self, destination, room_id, event_id, content):
+    async def send_invite_v1(self, destination, room_id, event_id, content):
         path = _create_v1_path("/invite/%s/%s", room_id, event_id)
 
-        response = yield self.client.put_json(
+        response = await self.client.put_json(
             destination=destination, path=path, data=content, ignore_backoff=True
         )
 
         return response
 
-    @defer.inlineCallbacks
     @log_function
-    def send_invite_v2(self, destination, room_id, event_id, content):
+    async def send_invite_v2(self, destination, room_id, event_id, content):
         path = _create_v2_path("/invite/%s/%s", room_id, event_id)
 
-        response = yield self.client.put_json(
+        response = await self.client.put_json(
             destination=destination, path=path, data=content, ignore_backoff=True
         )
 
         return response
 
-    @defer.inlineCallbacks
     @log_function
-    def get_public_rooms(
+    async def get_public_rooms(
         self,
         remote_server: str,
         limit: Optional[int] = None,
@@ -355,7 +345,7 @@ class TransportLayerClient(object):
             data["filter"] = search_filter
 
             try:
-                response = yield self.client.post_json(
+                response = await self.client.post_json(
                     destination=remote_server, path=path, data=data, ignore_backoff=True
                 )
             except HttpResponseException as e:
@@ -381,7 +371,7 @@ class TransportLayerClient(object):
                 args["since"] = [since_token]
 
             try:
-                response = yield self.client.get_json(
+                response = await self.client.get_json(
                     destination=remote_server, path=path, args=args, ignore_backoff=True
                 )
             except HttpResponseException as e:
@@ -396,29 +386,26 @@ class TransportLayerClient(object):
 
         return response
 
-    @defer.inlineCallbacks
     @log_function
-    def exchange_third_party_invite(self, destination, room_id, event_dict):
+    async def exchange_third_party_invite(self, destination, room_id, event_dict):
         path = _create_v1_path("/exchange_third_party_invite/%s", room_id)
 
-        response = yield self.client.put_json(
+        response = await self.client.put_json(
             destination=destination, path=path, data=event_dict
         )
 
         return response
 
-    @defer.inlineCallbacks
     @log_function
-    def get_event_auth(self, destination, room_id, event_id):
+    async def get_event_auth(self, destination, room_id, event_id):
         path = _create_v1_path("/event_auth/%s/%s", room_id, event_id)
 
-        content = yield self.client.get_json(destination=destination, path=path)
+        content = await self.client.get_json(destination=destination, path=path)
 
         return content
 
-    @defer.inlineCallbacks
     @log_function
-    def query_client_keys(self, destination, query_content, timeout):
+    async def query_client_keys(self, destination, query_content, timeout):
         """Query the device keys for a list of user ids hosted on a remote
         server.
 
@@ -453,14 +440,13 @@ class TransportLayerClient(object):
         """
         path = _create_v1_path("/user/keys/query")
 
-        content = yield self.client.post_json(
+        content = await self.client.post_json(
             destination=destination, path=path, data=query_content, timeout=timeout
         )
         return content
 
-    @defer.inlineCallbacks
     @log_function
-    def query_user_devices(self, destination, user_id, timeout):
+    async def query_user_devices(self, destination, user_id, timeout):
         """Query the devices for a user id hosted on a remote server.
 
         Response:
@@ -493,14 +479,13 @@ class TransportLayerClient(object):
         """
         path = _create_v1_path("/user/devices/%s", user_id)
 
-        content = yield self.client.get_json(
+        content = await self.client.get_json(
             destination=destination, path=path, timeout=timeout
         )
         return content
 
-    @defer.inlineCallbacks
     @log_function
-    def claim_client_keys(self, destination, query_content, timeout):
+    async def claim_client_keys(self, destination, query_content, timeout):
         """Claim one-time keys for a list of devices hosted on a remote server.
 
         Request:
@@ -532,14 +517,13 @@ class TransportLayerClient(object):
 
         path = _create_v1_path("/user/keys/claim")
 
-        content = yield self.client.post_json(
+        content = await self.client.post_json(
             destination=destination, path=path, data=query_content, timeout=timeout
         )
         return content
 
-    @defer.inlineCallbacks
     @log_function
-    def get_missing_events(
+    async def get_missing_events(
         self,
         destination,
         room_id,
@@ -551,7 +535,7 @@ class TransportLayerClient(object):
     ):
         path = _create_v1_path("/get_missing_events/%s", room_id)
 
-        content = yield self.client.post_json(
+        content = await self.client.post_json(
             destination=destination,
             path=path,
             data={

synapse/groups/attestations.py

@@ -41,8 +41,6 @@ from typing import Tuple
 
 from signedjson.sign import sign_json
 
-from twisted.internet import defer
-
 from synapse.api.errors import HttpResponseException, RequestSendFailed, SynapseError
 from synapse.metrics.background_process_metrics import run_as_background_process
 from synapse.types import get_domain_from_id
@@ -72,8 +70,9 @@ class GroupAttestationSigning(object):
         self.server_name = hs.hostname
         self.signing_key = hs.signing_key
 
-    @defer.inlineCallbacks
-    def verify_attestation(self, attestation, group_id, user_id, server_name=None):
+    async def verify_attestation(
+        self, attestation, group_id, user_id, server_name=None
+    ):
         """Verifies that the given attestation matches the given parameters.
 
         An optional server_name can be supplied to explicitly set which server's
@@ -102,7 +101,7 @@ class GroupAttestationSigning(object):
         if valid_until_ms < now:
             raise SynapseError(400, "Attestation expired")
 
-        yield self.keyring.verify_json_for_server(
+        await self.keyring.verify_json_for_server(
             server_name, attestation, now, "Group attestation"
         )
 
@@ -142,8 +141,7 @@ class GroupAttestionRenewer(object):
                 self._start_renew_attestations, 30 * 60 * 1000
             )
 
-    @defer.inlineCallbacks
-    def on_renew_attestation(self, group_id, user_id, content):
+    async def on_renew_attestation(self, group_id, user_id, content):
         """When a remote updates an attestation
         """
         attestation = content["attestation"]
@@ -151,11 +149,11 @@ class GroupAttestionRenewer(object):
         if not self.is_mine_id(group_id) and not self.is_mine_id(user_id):
             raise SynapseError(400, "Neither user not group are on this server")
 
-        yield self.attestations.verify_attestation(
+        await self.attestations.verify_attestation(
             attestation, user_id=user_id, group_id=group_id
         )
 
-        yield self.store.update_remote_attestion(group_id, user_id, attestation)
+        await self.store.update_remote_attestion(group_id, user_id, attestation)
 
         return {}
 
@@ -172,8 +170,7 @@ class GroupAttestionRenewer(object):
             now + UPDATE_ATTESTATION_TIME_MS
         )
 
-        @defer.inlineCallbacks
-        def _renew_attestation(group_user: Tuple[str, str]):
+        async def _renew_attestation(group_user: Tuple[str, str]):
             group_id, user_id = group_user
             try:
                 if not self.is_mine_id(group_id):
@@ -186,16 +183,16 @@ class GroupAttestionRenewer(object):
                         user_id,
                         group_id,
                     )
-                    yield self.store.remove_attestation_renewal(group_id, user_id)
+                    await self.store.remove_attestation_renewal(group_id, user_id)
                     return
 
                 attestation = self.attestations.create_attestation(group_id, user_id)
 
-                yield self.transport_client.renew_group_attestation(
+                await self.transport_client.renew_group_attestation(
                     destination, group_id, user_id, content={"attestation": attestation}
                 )
 
-                yield self.store.update_attestation_renewal(
+                await self.store.update_attestation_renewal(
                     group_id, user_id, attestation
                 )
             except (RequestSendFailed, HttpResponseException) as e:

synapse/handlers/acme.py

@@ -17,7 +17,6 @@ import logging
 
 import twisted
 import twisted.internet.error
-from twisted.internet import defer
 from twisted.web import server, static
 from twisted.web.resource import Resource
 
@@ -41,8 +40,7 @@ class AcmeHandler(object):
         self.reactor = hs.get_reactor()
         self._acme_domain = hs.config.acme_domain
 
-    @defer.inlineCallbacks
-    def start_listening(self):
+    async def start_listening(self):
         from synapse.handlers import acme_issuing_service
 
         # Configure logging for txacme, if you need to debug
@@ -82,18 +80,17 @@ class AcmeHandler(object):
         self._issuer._registered = False
 
         try:
-            yield self._issuer._ensure_registered()
+            await self._issuer._ensure_registered()
         except Exception:
             logger.error(ACME_REGISTER_FAIL_ERROR)
             raise
 
-    @defer.inlineCallbacks
-    def provision_certificate(self):
+    async def provision_certificate(self):
 
         logger.warning("Reprovisioning %s", self._acme_domain)
 
         try:
-            yield self._issuer.issue_cert(self._acme_domain)
+            await self._issuer.issue_cert(self._acme_domain)
         except Exception:
             logger.exception("Fail!")
             raise

synapse/handlers/appservice.py

@@ -27,7 +27,6 @@ from synapse.metrics import (
     event_processing_loop_room_count,
 )
 from synapse.metrics.background_process_metrics import run_as_background_process
-from synapse.util import log_failure
 from synapse.util.metrics import Measure
 
 logger = logging.getLogger(__name__)
@@ -100,10 +99,11 @@ class ApplicationServicesHandler(object):
 
                         if not self.started_scheduler:
 
-                            def start_scheduler():
-                                return self.scheduler.start().addErrback(
-                                    log_failure, "Application Services Failure"
-                                )
+                            async def start_scheduler():
+                                try:
+                                    return self.scheduler.start()
+                                except Exception:
+                                    logger.error("Application Services Failure")
 
                             run_as_background_process("as_scheduler", start_scheduler)
                             self.started_scheduler = True

synapse/handlers/federation.py

@@ -2242,8 +2242,6 @@ class FederationHandler(BaseHandler):
                 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:
             updated context object
         """
@@ -2251,7 +2249,7 @@ class FederationHandler(BaseHandler):
         room_version_obj = KNOWN_ROOM_VERSIONS[room_version]
 
         try:
-            context = await self._update_auth_events_and_context_for_auth(
+            context = await self._fetch_missing_auth_events(
                 origin, event, context, auth_events
             )
         except Exception:
@@ -2273,7 +2271,7 @@ class FederationHandler(BaseHandler):
 
         return context
 
-    async def _update_auth_events_and_context_for_auth(
+    async def _fetch_missing_auth_events(
         self,
         origin: str,
         event: EventBase,
@@ -2282,14 +2280,8 @@ class FederationHandler(BaseHandler):
     ) -> EventContext:
         """Helper for do_auth. See there for docs.
 
-        Checks whether a given event has the expected auth events. If it
-        doesn't then we talk to the remote server to compare state to see if
-        we can come to a consensus (e.g. if one server missed some valid
-        state).
-
-        This attempts to resolve any potential divergence of state between
-        servers, but is not essential and so failures should not block further
-        processing of the event.
+        Checks and fetches if there are any auth events that we don't have,
+        and if so fetch them.
 
         Args:
             origin:
@@ -2304,8 +2296,6 @@ class FederationHandler(BaseHandler):
                 event is an outlier), may be the auth events claimed by the remote
                 server.
 
-                Also NB that this function adds entries to it.
-
         Returns:
             updated context
         """
@@ -2363,141 +2353,14 @@ class FederationHandler(BaseHandler):
                             "do_auth %s missing_auth: %s", event.event_id, e.event_id
                         )
                         await self._handle_new_event(origin, e, auth_events=auth)
-
-                        if e.event_id in event_auth_events:
-                            auth_events[(e.type, e.state_key)] = e
                     except AuthError:
                         pass
 
             except Exception:
                 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
-
-        different_auth = event_auth_events.difference(
-            e.event_id for e in auth_events.values()
-        )
-
-        if not different_auth:
-            return context
-
-        logger.info(
-            "auth_events refers to events which are not in our calculated auth "
-            "chain: %s",
-            different_auth,
-        )
-
-        # XXX: currently this checks for redactions but I'm not convinced that is
-        # necessary?
-        different_events = await self.store.get_events_as_list(different_auth)
-
-        for d in different_events:
-            if d.room_id != event.room_id:
-                logger.warning(
-                    "Event %s refers to auth_event %s which is in a different room",
-                    event.event_id,
-                    d.event_id,
-                )
-
-                # don't attempt to resolve the claimed auth events against our own
-                # in this case: just use our own auth events.
-                #
-                # XXX: should we reject the event in this case? It feels like we should,
-                # but then shouldn't we also do so if we've failed to fetch any of the
-                # auth events?
-                return context
-
-        # now we state-resolve between our own idea of the auth events, and the remote's
-        # idea of them.
-
-        local_state = auth_events.values()
-        remote_auth_events = dict(auth_events)
-        remote_auth_events.update({(d.type, d.state_key): d for d in different_events})
-        remote_state = remote_auth_events.values()
-
-        room_version = await self.store.get_room_version_id(event.room_id)
-        new_state = await self.state_handler.resolve_events(
-            room_version, (local_state, remote_state), event
-        )
-
-        logger.info(
-            "After state res: updating auth_events with new state %s",
-            {
-                (d.type, d.state_key): d.event_id
-                for d in new_state.values()
-                if auth_events.get((d.type, d.state_key)) != d
-            },
-        )
-
-        auth_events.update(new_state)
-
-        context = await self._update_context_for_auth_events(
-            event, context, auth_events
-        )
-
         return context
 
-    async def _update_context_for_auth_events(
-        self, event: EventBase, context: EventContext, auth_events: StateMap[EventBase]
-    ) -> EventContext:
-        """Update the state_ids in an event context after auth event resolution,
-        storing the changes as a new state group.
-
-        Args:
-            event: The event we're handling the context for
-
-            context: initial event context
-
-            auth_events: Events to update in the event context.
-
-        Returns:
-            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)  # type: Optional[Tuple[str, str]]
-        else:
-            event_key = None
-        state_updates = {
-            k: a.event_id for k, a in auth_events.items() if k != event_key
-        }
-
-        current_state_ids = await context.get_current_state_ids()
-        current_state_ids = dict(current_state_ids)
-
-        current_state_ids.update(state_updates)
-
-        prev_state_ids = await context.get_prev_state_ids()
-        prev_state_ids = dict(prev_state_ids)
-
-        prev_state_ids.update({k: a.event_id for k, a in auth_events.items()})
-
-        # create a new state group as a delta from the existing one.
-        prev_group = context.state_group
-        state_group = await self.state_store.store_state_group(
-            event.event_id,
-            event.room_id,
-            prev_group=prev_group,
-            delta_ids=state_updates,
-            current_state_ids=current_state_ids,
-        )
-
-        return EventContext.with_state(
-            state_group=state_group,
-            state_group_before_event=context.state_group_before_event,
-            current_state_ids=current_state_ids,
-            prev_state_ids=prev_state_ids,
-            prev_group=prev_group,
-            delta_ids=state_updates,
-        )
-
     async def construct_auth_difference(
         self, local_auth: Iterable[EventBase], remote_auth: Iterable[EventBase]
     ) -> Dict:

synapse/handlers/groups_local.py

@@ -23,39 +23,32 @@ logger = logging.getLogger(__name__)
 
 
 def _create_rerouter(func_name):
-    """Returns a function that looks at the group id and calls the function
+    """Returns an async function that looks at the group id and calls the function
     on federation or the local group server if the group is local
     """
 
-    def f(self, group_id, *args, **kwargs):
+    async def f(self, group_id, *args, **kwargs):
         if self.is_mine_id(group_id):
-            return getattr(self.groups_server_handler, func_name)(
+            return await getattr(self.groups_server_handler, func_name)(
                 group_id, *args, **kwargs
             )
         else:
             destination = get_domain_from_id(group_id)
-            d = getattr(self.transport_client, func_name)(
-                destination, group_id, *args, **kwargs
-            )
 
-            # Capture errors returned by the remote homeserver and
-            # re-throw specific errors as SynapseErrors. This is so
-            # when the remote end responds with things like 403 Not
-            # In Group, we can communicate that to the client instead
-            # of a 500.
-            def http_response_errback(failure):
-                failure.trap(HttpResponseException)
-                e = failure.value
+            try:
+                return await getattr(self.transport_client, func_name)(
+                    destination, group_id, *args, **kwargs
+                )
+            except HttpResponseException as e:
+                # Capture errors returned by the remote homeserver and
+                # re-throw specific errors as SynapseErrors. This is so
+                # when the remote end responds with things like 403 Not
+                # In Group, we can communicate that to the client instead
+                # of a 500.
                 raise e.to_synapse_error()
-
-            def request_failed_errback(failure):
-                failure.trap(RequestSendFailed)
+            except RequestSendFailed:
                 raise SynapseError(502, "Failed to contact group server")
 
-            d.addErrback(http_response_errback)
-            d.addErrback(request_failed_errback)
-            return d
-
     return f
 
 

synapse/handlers/identity.py

@@ -22,14 +22,10 @@ import urllib.parse
 from typing import Awaitable, Callable, Dict, List, Optional, Tuple
 
 from canonicaljson import json
-from signedjson.key import decode_verify_key_bytes
-from signedjson.sign import verify_signed_json
-from unpaddedbase64 import decode_base64
 
 from twisted.internet.error import TimeoutError
 
 from synapse.api.errors import (
-    AuthError,
     CodeMessageException,
     Codes,
     HttpResponseException,
@@ -628,9 +624,9 @@ class IdentityHandler(BaseHandler):
             )
 
             if "mxid" in data:
-                if "signatures" not in data:
-                    raise AuthError(401, "No signatures on 3pid binding")
-                await self._verify_any_signature(data, id_server)
+                # note: we used to verify the identity server's signature here, but no longer
+                # require or validate it. See the following for context:
+                # https://github.com/matrix-org/synapse/issues/5253#issuecomment-666246950
                 return data["mxid"]
         except TimeoutError:
             raise SynapseError(500, "Timed out contacting identity server")
@@ -751,30 +747,6 @@ class IdentityHandler(BaseHandler):
         mxid = lookup_results["mappings"].get(lookup_value)
         return mxid
 
-    async def _verify_any_signature(self, data, server_hostname):
-        if server_hostname not in data["signatures"]:
-            raise AuthError(401, "No signature from server %s" % (server_hostname,))
-        for key_name, signature in data["signatures"][server_hostname].items():
-            try:
-                key_data = await self.blacklisting_http_client.get_json(
-                    "%s%s/_matrix/identity/api/v1/pubkey/%s"
-                    % (id_server_scheme, server_hostname, key_name)
-                )
-            except TimeoutError:
-                raise SynapseError(500, "Timed out contacting identity server")
-            if "public_key" not in key_data:
-                raise AuthError(
-                    401, "No public key named %s from %s" % (key_name, server_hostname)
-                )
-            verify_signed_json(
-                data,
-                server_hostname,
-                decode_verify_key_bytes(
-                    key_name, decode_base64(key_data["public_key"])
-                ),
-            )
-            return
-
     async def ask_id_server_for_third_party_invite(
         self,
         requester: Requester,

synapse/handlers/register.py

@@ -548,7 +548,7 @@ class RegistrationHandler(BaseHandler):
             address (str|None): the IP address used to perform the registration.
 
         Returns:
-            Deferred
+            Awaitable
         """
         if self.hs.config.worker_app:
             return self._register_client(

synapse/handlers/room_member.py

@@ -22,7 +22,8 @@ from unpaddedbase64 import encode_base64
 
 from synapse import types
 from synapse.api.constants import MAX_DEPTH, EventTypes, Membership
-from synapse.api.errors import AuthError, Codes, SynapseError
+from synapse.api.errors import AuthError, Codes, LimitExceededError, SynapseError
+from synapse.api.ratelimiting import Ratelimiter
 from synapse.api.room_versions import EventFormatVersions
 from synapse.crypto.event_signing import compute_event_reference_hash
 from synapse.events import EventBase
@@ -77,6 +78,17 @@ class RoomMemberHandler(object):
         if self._is_on_event_persistence_instance:
             self.persist_event_storage = hs.get_storage().persistence
 
+        self._join_rate_limiter_local = Ratelimiter(
+            clock=self.clock,
+            rate_hz=hs.config.ratelimiting.rc_joins_local.per_second,
+            burst_count=hs.config.ratelimiting.rc_joins_local.burst_count,
+        )
+        self._join_rate_limiter_remote = Ratelimiter(
+            clock=self.clock,
+            rate_hz=hs.config.ratelimiting.rc_joins_remote.per_second,
+            burst_count=hs.config.ratelimiting.rc_joins_remote.burst_count,
+        )
+
         # This is only used to get at ratelimit function, and
         # maybe_kick_guest_users. It's fine there are multiple of these as
         # it doesn't store state.
@@ -441,7 +453,28 @@ class RoomMemberHandler(object):
                     # so don't really fit into the general auth process.
                     raise AuthError(403, "Guest access not allowed")
 
-            if not is_host_in_room:
+            if is_host_in_room:
+                time_now_s = self.clock.time()
+                allowed, time_allowed = self._join_rate_limiter_local.can_do_action(
+                    requester.user.to_string(),
+                )
+
+                if not allowed:
+                    raise LimitExceededError(
+                        retry_after_ms=int(1000 * (time_allowed - time_now_s))
+                    )
+
+            else:
+                time_now_s = self.clock.time()
+                allowed, time_allowed = self._join_rate_limiter_remote.can_do_action(
+                    requester.user.to_string(),
+                )
+
+                if not allowed:
+                    raise LimitExceededError(
+                        retry_after_ms=int(1000 * (time_allowed - time_now_s))
+                    )
+
                 inviter = await self._get_inviter(target.to_string(), room_id)
                 if inviter and not self.hs.is_mine(inviter):
                     remote_room_hosts.append(inviter.domain)
@@ -469,26 +502,39 @@ class RoomMemberHandler(object):
                     user_id=target.to_string(), room_id=room_id
                 )  # type: Optional[RoomsForUser]
                 if not invite:
+                    logger.info(
+                        "%s sent a leave request to %s, but that is not an active room "
+                        "on this server, and there is no pending invite",
+                        target,
+                        room_id,
+                    )
+
                     raise SynapseError(404, "Not a known room")
 
                 logger.info(
                     "%s rejects invite to %s from %s", target, room_id, invite.sender
                 )
 
-                if self.hs.is_mine_id(invite.sender):
-                    # the inviter was on our server, but has now left. Carry on
-                    # with the normal rejection codepath.
-                    #
-                    # This is a bit of a hack, because the room might still be
-                    # active on other servers.
-                    pass
-                else:
+                if not self.hs.is_mine_id(invite.sender):
                     # send the rejection to the inviter's HS (with fallback to
                     # local event)
                     return await self.remote_reject_invite(
                         invite.event_id, txn_id, requester, content,
                     )
 
+                # the inviter was on our server, but has now left. Carry on
+                # with the normal rejection codepath, which will also send the
+                # rejection out to any other servers we believe are still in the room.
+
+                # thanks to overzealous cleaning up of event_forward_extremities in
+                # `delete_old_current_state_events`, it's possible to end up with no
+                # forward extremities here. If that happens, let's just hang the
+                # rejection off the invite event.
+                #
+                # see: https://github.com/matrix-org/synapse/issues/7139
+                if len(latest_event_ids) == 0:
+                    latest_event_ids = [invite.event_id]
+
         return await self._local_membership_update(
             requester=requester,
             target=target,
@@ -952,7 +998,11 @@ class RoomMemberMasterHandler(RoomMemberHandler):
         if len(remote_room_hosts) == 0:
             raise SynapseError(404, "No known servers")
 
-        if self.hs.config.limit_remote_rooms.enabled:
+        check_complexity = self.hs.config.limit_remote_rooms.enabled
+        if check_complexity and self.hs.config.limit_remote_rooms.admins_can_join:
+            check_complexity = not await self.hs.auth.is_server_admin(user)
+
+        if check_complexity:
             # Fetch the room complexity
             too_complex = await self._is_remote_room_too_complex(
                 room_id, remote_room_hosts
@@ -975,7 +1025,7 @@ class RoomMemberMasterHandler(RoomMemberHandler):
 
         # Check the room we just joined wasn't too large, if we didn't fetch the
         # complexity of it before.
-        if self.hs.config.limit_remote_rooms.enabled:
+        if check_complexity:
             if too_complex is False:
                 # We checked, and we're under the limit.
                 return event_id, stream_id

synapse/handlers/saml_handler.py

@@ -96,6 +96,9 @@ class SamlHandler:
             relay_state=client_redirect_url
         )
 
+        # Since SAML sessions timeout it is useful to log when they were created.
+        logger.info("Initiating a new SAML session: %s" % (reqid,))
+
         now = self._clock.time_msec()
         self._outstanding_requests_dict[reqid] = Saml2SessionData(
             creation_time=now, ui_auth_session_id=ui_auth_session_id,

synapse/handlers/stats.py

@@ -232,7 +232,7 @@ class StatsHandler:
 
                 if membership == prev_membership:
                     pass  # noop
-                if membership == Membership.JOIN:
+                elif membership == Membership.JOIN:
                     room_stats_delta["joined_members"] += 1
                 elif membership == Membership.INVITE:
                     room_stats_delta["invited_members"] += 1

synapse/handlers/sync.py

@@ -103,6 +103,7 @@ class JoinedSyncResult:
     account_data = attr.ib(type=List[JsonDict])
     unread_notifications = attr.ib(type=JsonDict)
     summary = attr.ib(type=Optional[JsonDict])
+    unread_count = attr.ib(type=int)
 
     def __nonzero__(self) -> bool:
         """Make the result appear empty if there are no updates. This is used
@@ -1886,6 +1887,10 @@ class SyncHandler(object):
 
         if room_builder.rtype == "joined":
             unread_notifications = {}  # type: Dict[str, str]
+
+            unread_count = await self.store.get_unread_message_count_for_user(
+                room_id, sync_config.user.to_string(),
+            )
             room_sync = JoinedSyncResult(
                 room_id=room_id,
                 timeline=batch,
@@ -1894,6 +1899,7 @@ class SyncHandler(object):
                 account_data=account_data_events,
                 unread_notifications=unread_notifications,
                 summary=summary,
+                unread_count=unread_count,
             )
 
             if room_sync or always_include:

synapse/http/client.py

@@ -395,7 +395,9 @@ class SimpleHttpClient(object):
         if 200 <= response.code < 300:
             return json.loads(body.decode("utf-8"))
         else:
-            raise HttpResponseException(response.code, response.phrase, body)
+            raise HttpResponseException(
+                response.code, response.phrase.decode("ascii", errors="replace"), body
+            )
 
     @defer.inlineCallbacks
     def post_json_get_json(self, uri, post_json, headers=None):
@@ -436,7 +438,9 @@ class SimpleHttpClient(object):
         if 200 <= response.code < 300:
             return json.loads(body.decode("utf-8"))
         else:
-            raise HttpResponseException(response.code, response.phrase, body)
+            raise HttpResponseException(
+                response.code, response.phrase.decode("ascii", errors="replace"), body
+            )
 
     @defer.inlineCallbacks
     def get_json(self, uri, args={}, headers=None):
@@ -509,7 +513,9 @@ class SimpleHttpClient(object):
         if 200 <= response.code < 300:
             return json.loads(body.decode("utf-8"))
         else:
-            raise HttpResponseException(response.code, response.phrase, body)
+            raise HttpResponseException(
+                response.code, response.phrase.decode("ascii", errors="replace"), body
+            )
 
     @defer.inlineCallbacks
     def get_raw(self, uri, args={}, headers=None):
@@ -544,7 +550,9 @@ class SimpleHttpClient(object):
         if 200 <= response.code < 300:
             return body
         else:
-            raise HttpResponseException(response.code, response.phrase, body)
+            raise HttpResponseException(
+                response.code, response.phrase.decode("ascii", errors="replace"), body
+            )
 
     # XXX: FIXME: This is horribly copy-pasted from matrixfederationclient.
     # The two should be factored out.

synapse/http/matrixfederationclient.py

@@ -121,8 +121,7 @@ class MatrixFederationRequest(object):
         return self.json
 
 
-@defer.inlineCallbacks
-def _handle_json_response(reactor, timeout_sec, request, response):
+async def _handle_json_response(reactor, timeout_sec, request, response):
     """
     Reads the JSON body of a response, with a timeout
 
@@ -141,7 +140,7 @@ def _handle_json_response(reactor, timeout_sec, request, response):
         d = treq.json_content(response)
         d = timeout_deferred(d, timeout=timeout_sec, reactor=reactor)
 
-        body = yield make_deferred_yieldable(d)
+        body = await make_deferred_yieldable(d)
     except TimeoutError as e:
         logger.warning(
             "{%s} [%s] Timed out reading response", request.txn_id, request.destination,
@@ -224,8 +223,7 @@ class MatrixFederationHttpClient(object):
 
         self._cooperator = Cooperator(scheduler=schedule)
 
-    @defer.inlineCallbacks
-    def _send_request_with_optional_trailing_slash(
+    async def _send_request_with_optional_trailing_slash(
         self, request, try_trailing_slash_on_400=False, **send_request_args
     ):
         """Wrapper for _send_request which can optionally retry the request
@@ -246,10 +244,10 @@ class MatrixFederationHttpClient(object):
                 (except 429).
 
         Returns:
-            Deferred[Dict]: Parsed JSON response body.
+            Dict: Parsed JSON response body.
         """
         try:
-            response = yield self._send_request(request, **send_request_args)
+            response = await self._send_request(request, **send_request_args)
         except HttpResponseException as e:
             # Received an HTTP error > 300. Check if it meets the requirements
             # to retry with a trailing slash
@@ -265,12 +263,11 @@ class MatrixFederationHttpClient(object):
             logger.info("Retrying request with trailing slash")
             request.path += "/"
 
-            response = yield self._send_request(request, **send_request_args)
+            response = await self._send_request(request, **send_request_args)
 
         return response
 
-    @defer.inlineCallbacks
-    def _send_request(
+    async def _send_request(
         self,
         request,
         retry_on_dns_fail=True,
@@ -311,7 +308,7 @@ class MatrixFederationHttpClient(object):
             backoff_on_404 (bool): Back off if we get a 404
 
         Returns:
-            Deferred[twisted.web.client.Response]: resolves with the HTTP
+            twisted.web.client.Response: resolves with the HTTP
             response object on success.
 
         Raises:
@@ -335,7 +332,7 @@ class MatrixFederationHttpClient(object):
         ):
             raise FederationDeniedError(request.destination)
 
-        limiter = yield synapse.util.retryutils.get_retry_limiter(
+        limiter = await synapse.util.retryutils.get_retry_limiter(
             request.destination,
             self.clock,
             self._store,
@@ -433,7 +430,7 @@ class MatrixFederationHttpClient(object):
                                 reactor=self.reactor,
                             )
 
-                            response = yield request_deferred
+                            response = await request_deferred
                     except TimeoutError as e:
                         raise RequestSendFailed(e, can_retry=True) from e
                     except DNSLookupError as e:
@@ -447,6 +444,7 @@ class MatrixFederationHttpClient(object):
                     ).inc()
 
                     set_tag(tags.HTTP_STATUS_CODE, response.code)
+                    response_phrase = response.phrase.decode("ascii", errors="replace")
 
                     if 200 <= response.code < 300:
                         logger.debug(
@@ -454,7 +452,7 @@ class MatrixFederationHttpClient(object):
                             request.txn_id,
                             request.destination,
                             response.code,
-                            response.phrase.decode("ascii", errors="replace"),
+                            response_phrase,
                         )
                         pass
                     else:
@@ -463,7 +461,7 @@ class MatrixFederationHttpClient(object):
                             request.txn_id,
                             request.destination,
                             response.code,
-                            response.phrase.decode("ascii", errors="replace"),
+                            response_phrase,
                         )
                         # :'(
                         # Update transactions table?
@@ -473,7 +471,7 @@ class MatrixFederationHttpClient(object):
                         )
 
                         try:
-                            body = yield make_deferred_yieldable(d)
+                            body = await make_deferred_yieldable(d)
                         except Exception as e:
                             # Eh, we're already going to raise an exception so lets
                             # ignore if this fails.
@@ -487,7 +485,7 @@ class MatrixFederationHttpClient(object):
                             )
                             body = None
 
-                        e = HttpResponseException(response.code, response.phrase, body)
+                        e = HttpResponseException(response.code, response_phrase, body)
 
                         # Retry if the error is a 429 (Too Many Requests),
                         # otherwise just raise a standard HttpResponseException
@@ -527,7 +525,7 @@ class MatrixFederationHttpClient(object):
                             delay,
                         )
 
-                        yield self.clock.sleep(delay)
+                        await self.clock.sleep(delay)
                         retries_left -= 1
                     else:
                         raise
@@ -590,8 +588,7 @@ class MatrixFederationHttpClient(object):
             )
         return auth_headers
 
-    @defer.inlineCallbacks
-    def put_json(
+    async def put_json(
         self,
         destination,
         path,
@@ -635,7 +632,7 @@ class MatrixFederationHttpClient(object):
                 enabled.
 
         Returns:
-            Deferred[dict|list]: Succeeds when we get a 2xx HTTP response. The
+            dict|list: Succeeds when we get a 2xx HTTP response. The
             result will be the decoded JSON body.
 
         Raises:
@@ -657,7 +654,7 @@ class MatrixFederationHttpClient(object):
             json=data,
         )
 
-        response = yield self._send_request_with_optional_trailing_slash(
+        response = await self._send_request_with_optional_trailing_slash(
             request,
             try_trailing_slash_on_400,
             backoff_on_404=backoff_on_404,
@@ -666,14 +663,13 @@ class MatrixFederationHttpClient(object):
             timeout=timeout,
         )
 
-        body = yield _handle_json_response(
+        body = await _handle_json_response(
             self.reactor, self.default_timeout, request, response
         )
 
         return body
 
-    @defer.inlineCallbacks
-    def post_json(
+    async def post_json(
         self,
         destination,
         path,
@@ -706,7 +702,7 @@ class MatrixFederationHttpClient(object):
 
             args (dict): query params
         Returns:
-            Deferred[dict|list]: Succeeds when we get a 2xx HTTP response. The
+            dict|list: Succeeds when we get a 2xx HTTP response. The
             result will be the decoded JSON body.
 
         Raises:
@@ -724,7 +720,7 @@ class MatrixFederationHttpClient(object):
             method="POST", destination=destination, path=path, query=args, json=data
         )
 
-        response = yield self._send_request(
+        response = await self._send_request(
             request,
             long_retries=long_retries,
             timeout=timeout,
@@ -736,13 +732,12 @@ class MatrixFederationHttpClient(object):
         else:
             _sec_timeout = self.default_timeout
 
-        body = yield _handle_json_response(
+        body = await _handle_json_response(
             self.reactor, _sec_timeout, request, response
         )
         return body
 
-    @defer.inlineCallbacks
-    def get_json(
+    async def get_json(
         self,
         destination,
         path,
@@ -774,7 +769,7 @@ class MatrixFederationHttpClient(object):
                 response we should try appending a trailing slash to the end of
                 the request. Workaround for #3622 in Synapse <= v0.99.3.
         Returns:
-            Deferred[dict|list]: Succeeds when we get a 2xx HTTP response. The
+            dict|list: Succeeds when we get a 2xx HTTP response. The
             result will be the decoded JSON body.
 
         Raises:
@@ -791,7 +786,7 @@ class MatrixFederationHttpClient(object):
             method="GET", destination=destination, path=path, query=args
         )
 
-        response = yield self._send_request_with_optional_trailing_slash(
+        response = await self._send_request_with_optional_trailing_slash(
             request,
             try_trailing_slash_on_400,
             backoff_on_404=False,
@@ -800,14 +795,13 @@ class MatrixFederationHttpClient(object):
             timeout=timeout,
         )
 
-        body = yield _handle_json_response(
+        body = await _handle_json_response(
             self.reactor, self.default_timeout, request, response
         )
 
         return body
 
-    @defer.inlineCallbacks
-    def delete_json(
+    async def delete_json(
         self,
         destination,
         path,
@@ -835,7 +829,7 @@ class MatrixFederationHttpClient(object):
 
             args (dict): query params
         Returns:
-            Deferred[dict|list]: Succeeds when we get a 2xx HTTP response. The
+            dict|list: Succeeds when we get a 2xx HTTP response. The
             result will be the decoded JSON body.
 
         Raises:
@@ -852,20 +846,19 @@ class MatrixFederationHttpClient(object):
             method="DELETE", destination=destination, path=path, query=args
         )
 
-        response = yield self._send_request(
+        response = await self._send_request(
             request,
             long_retries=long_retries,
             timeout=timeout,
             ignore_backoff=ignore_backoff,
         )
 
-        body = yield _handle_json_response(
+        body = await _handle_json_response(
             self.reactor, self.default_timeout, request, response
         )
         return body
 
-    @defer.inlineCallbacks
-    def get_file(
+    async def get_file(
         self,
         destination,
         path,
@@ -885,7 +878,7 @@ class MatrixFederationHttpClient(object):
                 and try the request anyway.
 
         Returns:
-            Deferred[tuple[int, dict]]: Resolves with an (int,dict) tuple of
+            tuple[int, dict]: Resolves with an (int,dict) tuple of
             the file length and a dict of the response headers.
 
         Raises:
@@ -902,7 +895,7 @@ class MatrixFederationHttpClient(object):
             method="GET", destination=destination, path=path, query=args
         )
 
-        response = yield self._send_request(
+        response = await self._send_request(
             request, retry_on_dns_fail=retry_on_dns_fail, ignore_backoff=ignore_backoff
         )
 
@@ -911,7 +904,7 @@ class MatrixFederationHttpClient(object):
         try:
             d = _readBodyToFile(response, output_stream, max_size)
             d.addTimeout(self.default_timeout, self.reactor)
-            length = yield make_deferred_yieldable(d)
+            length = await make_deferred_yieldable(d)
         except Exception as e:
             logger.warning(
                 "{%s} [%s] Error reading response: %s",

synapse/http/server.py

@@ -242,10 +242,12 @@ class _AsyncResource(resource.Resource, metaclass=abc.ABCMeta):
         no appropriate method exists. Can be overriden in sub classes for
         different routing.
         """
+        # Treat HEAD requests as GET requests.
+        request_method = request.method.decode("ascii")
+        if request_method == "HEAD":
+            request_method = "GET"
 
-        method_handler = getattr(
-            self, "_async_render_%s" % (request.method.decode("ascii"),), None
-        )
+        method_handler = getattr(self, "_async_render_%s" % (request_method,), None)
         if method_handler:
             raw_callback_return = method_handler(request)
 
@@ -362,11 +364,15 @@ class JsonResource(DirectServeJsonResource):
             A tuple of the callback to use, the name of the servlet, and the
             key word arguments to pass to the callback
         """
+        # Treat HEAD requests as GET requests.
         request_path = request.path.decode("ascii")
+        request_method = request.method
+        if request_method == b"HEAD":
+            request_method = b"GET"
 
         # Loop through all the registered callbacks to check if the method
         # and path regex match
-        for path_entry in self.path_regexs.get(request.method, []):
+        for path_entry in self.path_regexs.get(request_method, []):
             m = path_entry.pattern.match(request_path)
             if m:
                 # We found a match!
@@ -579,7 +585,7 @@ def set_cors_headers(request: Request):
     """
     request.setHeader(b"Access-Control-Allow-Origin", b"*")
     request.setHeader(
-        b"Access-Control-Allow-Methods", b"GET, POST, PUT, DELETE, OPTIONS"
+        b"Access-Control-Allow-Methods", b"GET, HEAD, POST, PUT, DELETE, OPTIONS"
     )
     request.setHeader(
         b"Access-Control-Allow-Headers",

synapse/push/action_generator.py

@@ -15,8 +15,6 @@
 
 import logging
 
-from twisted.internet import defer
-
 from synapse.util.metrics import Measure
 
 from .bulk_push_rule_evaluator import BulkPushRuleEvaluator
@@ -37,7 +35,6 @@ class ActionGenerator(object):
         # event stream, so we just run the rules for a client with no profile
         # tag (ie. we just need all the users).
 
-    @defer.inlineCallbacks
-    def handle_push_actions_for_event(self, event, context):
+    async def handle_push_actions_for_event(self, event, context):
         with Measure(self.clock, "action_for_event_by_user"):
-            yield self.bulk_evaluator.action_for_event_by_user(event, context)
+            await self.bulk_evaluator.action_for_event_by_user(event, context)

synapse/push/bulk_push_rule_evaluator.py

@@ -19,8 +19,6 @@ from collections import namedtuple
 
 from prometheus_client import Counter
 
-from twisted.internet import defer
-
 from synapse.api.constants import EventTypes, Membership
 from synapse.event_auth import get_user_power_level
 from synapse.state import POWER_KEY
@@ -70,8 +68,7 @@ class BulkPushRuleEvaluator(object):
             resizable=False,
         )
 
-    @defer.inlineCallbacks
-    def _get_rules_for_event(self, event, context):
+    async def _get_rules_for_event(self, event, context):
         """This gets the rules for all users in the room at the time of the event,
         as well as the push rules for the invitee if the event is an invite.
 
@@ -79,19 +76,19 @@ class BulkPushRuleEvaluator(object):
             dict of user_id -> push_rules
         """
         room_id = event.room_id
-        rules_for_room = yield self._get_rules_for_room(room_id)
+        rules_for_room = await self._get_rules_for_room(room_id)
 
-        rules_by_user = yield rules_for_room.get_rules(event, context)
+        rules_by_user = await rules_for_room.get_rules(event, context)
 
         # if this event is an invite event, we may need to run rules for the user
         # who's been invited, otherwise they won't get told they've been invited
         if event.type == "m.room.member" and event.content["membership"] == "invite":
             invited = event.state_key
             if invited and self.hs.is_mine_id(invited):
-                has_pusher = yield self.store.user_has_pusher(invited)
+                has_pusher = await self.store.user_has_pusher(invited)
                 if has_pusher:
                     rules_by_user = dict(rules_by_user)
-                    rules_by_user[invited] = yield self.store.get_push_rules_for_user(
+                    rules_by_user[invited] = await self.store.get_push_rules_for_user(
                         invited
                     )
 
@@ -114,20 +111,19 @@ class BulkPushRuleEvaluator(object):
             self.room_push_rule_cache_metrics,
         )
 
-    @defer.inlineCallbacks
-    def _get_power_levels_and_sender_level(self, event, context):
-        prev_state_ids = yield context.get_prev_state_ids()
+    async def _get_power_levels_and_sender_level(self, event, context):
+        prev_state_ids = await context.get_prev_state_ids()
         pl_event_id = prev_state_ids.get(POWER_KEY)
         if pl_event_id:
             # fastpath: if there's a power level event, that's all we need, and
             # not having a power level event is an extreme edge case
-            pl_event = yield self.store.get_event(pl_event_id)
+            pl_event = await self.store.get_event(pl_event_id)
             auth_events = {POWER_KEY: pl_event}
         else:
-            auth_events_ids = yield self.auth.compute_auth_events(
+            auth_events_ids = await self.auth.compute_auth_events(
                 event, prev_state_ids, for_verification=False
             )
-            auth_events = yield self.store.get_events(auth_events_ids)
+            auth_events = await self.store.get_events(auth_events_ids)
             auth_events = {(e.type, e.state_key): e for e in auth_events.values()}
 
         sender_level = get_user_power_level(event.sender, auth_events)
@@ -136,23 +132,19 @@ class BulkPushRuleEvaluator(object):
 
         return pl_event.content if pl_event else {}, sender_level
 
-    @defer.inlineCallbacks
-    def action_for_event_by_user(self, event, context):
+    async def action_for_event_by_user(self, event, context) -> None:
         """Given an event and context, evaluate the push rules and insert the
         results into the event_push_actions_staging table.
-
-        Returns:
-            Deferred
         """
-        rules_by_user = yield self._get_rules_for_event(event, context)
+        rules_by_user = await self._get_rules_for_event(event, context)
         actions_by_user = {}
 
-        room_members = yield self.store.get_joined_users_from_context(event, context)
+        room_members = await self.store.get_joined_users_from_context(event, context)
 
         (
             power_levels,
             sender_power_level,
-        ) = yield self._get_power_levels_and_sender_level(event, context)
+        ) = await self._get_power_levels_and_sender_level(event, context)
 
         evaluator = PushRuleEvaluatorForEvent(
             event, len(room_members), sender_power_level, power_levels
@@ -165,7 +157,7 @@ class BulkPushRuleEvaluator(object):
                 continue
 
             if not event.is_state():
-                is_ignored = yield self.store.is_ignored_by(event.sender, uid)
+                is_ignored = await self.store.is_ignored_by(event.sender, uid)
                 if is_ignored:
                     continue
 
@@ -197,7 +189,7 @@ class BulkPushRuleEvaluator(object):
         # Mark in the DB staging area the push actions for users who should be
         # notified for this event. (This will then get handled when we persist
         # the event)
-        yield self.store.add_push_actions_to_staging(event.event_id, actions_by_user)
+        await self.store.add_push_actions_to_staging(event.event_id, actions_by_user)
 
 
 def _condition_checker(evaluator, conditions, uid, display_name, cache):
@@ -274,8 +266,7 @@ class RulesForRoom(object):
         # to self around in the callback.
         self.invalidate_all_cb = _Invalidation(rules_for_room_cache, room_id)
 
-    @defer.inlineCallbacks
-    def get_rules(self, event, context):
+    async def get_rules(self, event, context):
         """Given an event context return the rules for all users who are
         currently in the room.
         """
@@ -286,7 +277,7 @@ class RulesForRoom(object):
             self.room_push_rule_cache_metrics.inc_hits()
             return self.rules_by_user
 
-        with (yield self.linearizer.queue(())):
+        with (await self.linearizer.queue(())):
             if state_group and self.state_group == state_group:
                 logger.debug("Using cached rules for %r", self.room_id)
                 self.room_push_rule_cache_metrics.inc_hits()
@@ -304,9 +295,7 @@ class RulesForRoom(object):
 
                 push_rules_delta_state_cache_metric.inc_hits()
             else:
-                current_state_ids = yield defer.ensureDeferred(
-                    context.get_current_state_ids()
-                )
+                current_state_ids = await context.get_current_state_ids()
                 push_rules_delta_state_cache_metric.inc_misses()
 
             push_rules_state_size_counter.inc(len(current_state_ids))
@@ -353,7 +342,7 @@ class RulesForRoom(object):
                 # If we have some memebr events we haven't seen, look them up
                 # and fetch push rules for them if appropriate.
                 logger.debug("Found new member events %r", missing_member_event_ids)
-                yield self._update_rules_with_member_event_ids(
+                await self._update_rules_with_member_event_ids(
                     ret_rules_by_user, missing_member_event_ids, state_group, event
                 )
             else:
@@ -371,8 +360,7 @@ class RulesForRoom(object):
             )
         return ret_rules_by_user
 
-    @defer.inlineCallbacks
-    def _update_rules_with_member_event_ids(
+    async def _update_rules_with_member_event_ids(
         self, ret_rules_by_user, member_event_ids, state_group, event
     ):
         """Update the partially filled rules_by_user dict by fetching rules for
@@ -388,7 +376,7 @@ class RulesForRoom(object):
         """
         sequence = self.sequence
 
-        rows = yield self.store.get_membership_from_event_ids(member_event_ids.values())
+        rows = await self.store.get_membership_from_event_ids(member_event_ids.values())
 
         members = {row["event_id"]: (row["user_id"], row["membership"]) for row in rows}
 
@@ -410,7 +398,7 @@ class RulesForRoom(object):
 
         logger.debug("Joined: %r", interested_in_user_ids)
 
-        if_users_with_pushers = yield self.store.get_if_users_have_pushers(
+        if_users_with_pushers = await self.store.get_if_users_have_pushers(
             interested_in_user_ids, on_invalidate=self.invalidate_all_cb
         )
 
@@ -420,7 +408,7 @@ class RulesForRoom(object):
 
         logger.debug("With pushers: %r", user_ids)
 
-        users_with_receipts = yield self.store.get_users_with_read_receipts_in_room(
+        users_with_receipts = await self.store.get_users_with_read_receipts_in_room(
             self.room_id, on_invalidate=self.invalidate_all_cb
         )
 
@@ -431,7 +419,7 @@ class RulesForRoom(object):
             if uid in interested_in_user_ids:
                 user_ids.add(uid)
 
-        rules_by_user = yield self.store.bulk_get_push_rules(
+        rules_by_user = await self.store.bulk_get_push_rules(
             user_ids, on_invalidate=self.invalidate_all_cb
         )
 

synapse/push/httppusher.py

@@ -17,7 +17,6 @@ import logging
 
 from prometheus_client import Counter
 
-from twisted.internet import defer
 from twisted.internet.error import AlreadyCalled, AlreadyCancelled
 
 from synapse.api.constants import EventTypes
@@ -128,12 +127,11 @@ class HttpPusher(object):
         # but currently that's the only type of receipt anyway...
         run_as_background_process("http_pusher.on_new_receipts", self._update_badge)
 
-    @defer.inlineCallbacks
-    def _update_badge(self):
+    async def _update_badge(self):
         # XXX as per https://github.com/matrix-org/matrix-doc/issues/2627, this seems
         # to be largely redundant. perhaps we can remove it.
-        badge = yield push_tools.get_badge_count(self.hs.get_datastore(), self.user_id)
-        yield self._send_badge(badge)
+        badge = await push_tools.get_badge_count(self.hs.get_datastore(), self.user_id)
+        await self._send_badge(badge)
 
     def on_timer(self):
         self._start_processing()
@@ -152,8 +150,7 @@ class HttpPusher(object):
 
         run_as_background_process("httppush.process", self._process)
 
-    @defer.inlineCallbacks
-    def _process(self):
+    async def _process(self):
         # we should never get here if we are already processing
         assert not self._is_processing
 
@@ -164,7 +161,7 @@ class HttpPusher(object):
             while True:
                 starting_max_ordering = self.max_stream_ordering
                 try:
-                    yield self._unsafe_process()
+                    await self._unsafe_process()
                 except Exception:
                     logger.exception("Exception processing notifs")
                 if self.max_stream_ordering == starting_max_ordering:
@@ -172,8 +169,7 @@ class HttpPusher(object):
         finally:
             self._is_processing = False
 
-    @defer.inlineCallbacks
-    def _unsafe_process(self):
+    async def _unsafe_process(self):
         """
         Looks for unset notifications and dispatch them, in order
         Never call this directly: use _process which will only allow this to
@@ -181,7 +177,7 @@ class HttpPusher(object):
         """
 
         fn = self.store.get_unread_push_actions_for_user_in_range_for_http
-        unprocessed = yield fn(
+        unprocessed = await fn(
             self.user_id, self.last_stream_ordering, self.max_stream_ordering
         )
 
@@ -203,13 +199,13 @@ class HttpPusher(object):
                     "app_display_name": self.app_display_name,
                 },
             ):
-                processed = yield self._process_one(push_action)
+                processed = await self._process_one(push_action)
 
             if processed:
                 http_push_processed_counter.inc()
                 self.backoff_delay = HttpPusher.INITIAL_BACKOFF_SEC
                 self.last_stream_ordering = push_action["stream_ordering"]
-                pusher_still_exists = yield self.store.update_pusher_last_stream_ordering_and_success(
+                pusher_still_exists = await self.store.update_pusher_last_stream_ordering_and_success(
                     self.app_id,
                     self.pushkey,
                     self.user_id,
@@ -224,14 +220,14 @@ class HttpPusher(object):
 
                 if self.failing_since:
                     self.failing_since = None
-                    yield self.store.update_pusher_failing_since(
+                    await self.store.update_pusher_failing_since(
                         self.app_id, self.pushkey, self.user_id, self.failing_since
                     )
             else:
                 http_push_failed_counter.inc()
                 if not self.failing_since:
                     self.failing_since = self.clock.time_msec()
-                    yield self.store.update_pusher_failing_since(
+                    await self.store.update_pusher_failing_since(
                         self.app_id, self.pushkey, self.user_id, self.failing_since
                     )
 
@@ -250,7 +246,7 @@ class HttpPusher(object):
                     )
                     self.backoff_delay = HttpPusher.INITIAL_BACKOFF_SEC
                     self.last_stream_ordering = push_action["stream_ordering"]
-                    pusher_still_exists = yield self.store.update_pusher_last_stream_ordering(
+                    pusher_still_exists = await self.store.update_pusher_last_stream_ordering(
                         self.app_id,
                         self.pushkey,
                         self.user_id,
@@ -263,7 +259,7 @@ class HttpPusher(object):
                         return
 
                     self.failing_since = None
-                    yield self.store.update_pusher_failing_since(
+                    await self.store.update_pusher_failing_since(
                         self.app_id, self.pushkey, self.user_id, self.failing_since
                     )
                 else:
@@ -276,18 +272,17 @@ class HttpPusher(object):
                     )
                     break
 
-    @defer.inlineCallbacks
-    def _process_one(self, push_action):
+    async def _process_one(self, push_action):
         if "notify" not in push_action["actions"]:
             return True
 
         tweaks = push_rule_evaluator.tweaks_for_actions(push_action["actions"])
-        badge = yield push_tools.get_badge_count(self.hs.get_datastore(), self.user_id)
+        badge = await push_tools.get_badge_count(self.hs.get_datastore(), self.user_id)
 
-        event = yield self.store.get_event(push_action["event_id"], allow_none=True)
+        event = await self.store.get_event(push_action["event_id"], allow_none=True)
         if event is None:
             return True  # It's been redacted
-        rejected = yield self.dispatch_push(event, tweaks, badge)
+        rejected = await self.dispatch_push(event, tweaks, badge)
         if rejected is False:
             return False
 
@@ -301,11 +296,10 @@ class HttpPusher(object):
                     )
                 else:
                     logger.info("Pushkey %s was rejected: removing", pk)
-                    yield self.hs.remove_pusher(self.app_id, pk, self.user_id)
+                    await self.hs.remove_pusher(self.app_id, pk, self.user_id)
         return True
 
-    @defer.inlineCallbacks
-    def _build_notification_dict(self, event, tweaks, badge):
+    async def _build_notification_dict(self, event, tweaks, badge):
         priority = "low"
         if (
             event.type == EventTypes.Encrypted
@@ -335,7 +329,7 @@ class HttpPusher(object):
             }
             return d
 
-        ctx = yield push_tools.get_context_for_event(
+        ctx = await push_tools.get_context_for_event(
             self.storage, self.state_handler, event, self.user_id
         )
 
@@ -377,13 +371,12 @@ class HttpPusher(object):
 
         return d
 
-    @defer.inlineCallbacks
-    def dispatch_push(self, event, tweaks, badge):
-        notification_dict = yield self._build_notification_dict(event, tweaks, badge)
+    async def dispatch_push(self, event, tweaks, badge):
+        notification_dict = await self._build_notification_dict(event, tweaks, badge)
         if not notification_dict:
             return []
         try:
-            resp = yield self.http_client.post_json_get_json(
+            resp = await self.http_client.post_json_get_json(
                 self.url, notification_dict
             )
         except Exception as e:
@@ -400,8 +393,7 @@ class HttpPusher(object):
             rejected = resp["rejected"]
         return rejected
 
-    @defer.inlineCallbacks
-    def _send_badge(self, badge):
+    async def _send_badge(self, badge):
         """
         Args:
             badge (int): number of unread messages
@@ -424,7 +416,7 @@ class HttpPusher(object):
             }
         }
         try:
-            yield self.http_client.post_json_get_json(self.url, d)
+            await self.http_client.post_json_get_json(self.url, d)
             http_badges_processed_counter.inc()
         except Exception as e:
             logger.warning(

synapse/push/presentable_names.py

@@ -16,8 +16,6 @@
 import logging
 import re
 
-from twisted.internet import defer
-
 from synapse.api.constants import EventTypes
 
 logger = logging.getLogger(__name__)
@@ -29,8 +27,7 @@ ALIAS_RE = re.compile(r"^#.*:.+$")
 ALL_ALONE = "Empty Room"
 
 
-@defer.inlineCallbacks
-def calculate_room_name(
+async def calculate_room_name(
     store,
     room_state_ids,
     user_id,
@@ -53,7 +50,7 @@ def calculate_room_name(
     """
     # does it have a name?
     if (EventTypes.Name, "") in room_state_ids:
-        m_room_name = yield store.get_event(
+        m_room_name = await store.get_event(
             room_state_ids[(EventTypes.Name, "")], allow_none=True
         )
         if m_room_name and m_room_name.content and m_room_name.content["name"]:
@@ -61,7 +58,7 @@ def calculate_room_name(
 
     # does it have a canonical alias?
     if (EventTypes.CanonicalAlias, "") in room_state_ids:
-        canon_alias = yield store.get_event(
+        canon_alias = await store.get_event(
             room_state_ids[(EventTypes.CanonicalAlias, "")], allow_none=True
         )
         if (
@@ -81,7 +78,7 @@ def calculate_room_name(
 
     my_member_event = None
     if (EventTypes.Member, user_id) in room_state_ids:
-        my_member_event = yield store.get_event(
+        my_member_event = await store.get_event(
             room_state_ids[(EventTypes.Member, user_id)], allow_none=True
         )
 
@@ -90,7 +87,7 @@ def calculate_room_name(
         and my_member_event.content["membership"] == "invite"
     ):
         if (EventTypes.Member, my_member_event.sender) in room_state_ids:
-            inviter_member_event = yield store.get_event(
+            inviter_member_event = await store.get_event(
                 room_state_ids[(EventTypes.Member, my_member_event.sender)],
                 allow_none=True,
             )
@@ -107,7 +104,7 @@ def calculate_room_name(
     # we're going to have to generate a name based on who's in the room,
     # so find out who is in the room that isn't the user.
     if EventTypes.Member in room_state_bytype_ids:
-        member_events = yield store.get_events(
+        member_events = await store.get_events(
             list(room_state_bytype_ids[EventTypes.Member].values())
         )
         all_members = [

synapse/push/push_tools.py

@@ -13,53 +13,40 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-from twisted.internet import defer
-
 from synapse.push.presentable_names import calculate_room_name, name_from_member_event
 from synapse.storage import Storage
 
 
-@defer.inlineCallbacks
-def get_badge_count(store, user_id):
-    invites = yield store.get_invited_rooms_for_local_user(user_id)
-    joins = yield store.get_rooms_for_user(user_id)
-
-    my_receipts_by_room = yield store.get_receipts_for_user(user_id, "m.read")
+async def get_badge_count(store, user_id):
+    invites = await store.get_invited_rooms_for_local_user(user_id)
+    joins = await store.get_rooms_for_user(user_id)
 
     badge = len(invites)
 
     for room_id in joins:
-        if room_id in my_receipts_by_room:
-            last_unread_event_id = my_receipts_by_room[room_id]
-
-            notifs = yield (
-                store.get_unread_event_push_actions_by_room_for_user(
-                    room_id, user_id, last_unread_event_id
-                )
-            )
-            # return one badge count per conversation, as count per
-            # message is so noisy as to be almost useless
-            badge += 1 if notifs["notify_count"] else 0
+        unread_count = await store.get_unread_message_count_for_user(room_id, user_id)
+        # return one badge count per conversation, as count per
+        # message is so noisy as to be almost useless
+        badge += 1 if unread_count else 0
     return badge
 
 
-@defer.inlineCallbacks
-def get_context_for_event(storage: Storage, state_handler, ev, user_id):
+async def get_context_for_event(storage: Storage, state_handler, ev, user_id):
     ctx = {}
 
-    room_state_ids = yield storage.state.get_state_ids_for_event(ev.event_id)
+    room_state_ids = await storage.state.get_state_ids_for_event(ev.event_id)
 
     # we no longer bother setting room_alias, and make room_name the
     # human-readable name instead, be that m.room.name, an alias or
     # a list of people in the room
-    name = yield calculate_room_name(
+    name = await calculate_room_name(
         storage.main, room_state_ids, user_id, fallback_to_single_member=False
     )
     if name:
         ctx["name"] = name
 
     sender_state_event_id = room_state_ids[("m.room.member", ev.sender)]
-    sender_state_event = yield storage.main.get_event(sender_state_event_id)
+    sender_state_event = await storage.main.get_event(sender_state_event_id)
     ctx["sender_display_name"] = name_from_member_event(sender_state_event)
 
     return ctx

synapse/push/pusherpool.py

@@ -19,8 +19,6 @@ from typing import TYPE_CHECKING, Dict, Union
 
 from prometheus_client import Gauge
 
-from twisted.internet import defer
-
 from synapse.metrics.background_process_metrics import run_as_background_process
 from synapse.push import PusherConfigException
 from synapse.push.emailpusher import EmailPusher
@@ -52,7 +50,7 @@ class PusherPool:
     Note that it is expected that each pusher will have its own 'processing' loop which
     will send out the notifications in the background, rather than blocking until the
     notifications are sent; accordingly Pusher.on_started, Pusher.on_new_notifications and
-    Pusher.on_new_receipts are not expected to return deferreds.
+    Pusher.on_new_receipts are not expected to return awaitables.
     """
 
     def __init__(self, hs: "HomeServer"):
@@ -77,8 +75,7 @@ class PusherPool:
             return
         run_as_background_process("start_pushers", self._start_pushers)
 
-    @defer.inlineCallbacks
-    def add_pusher(
+    async def add_pusher(
         self,
         user_id,
         access_token,
@@ -94,7 +91,7 @@ class PusherPool:
         """Creates a new pusher and adds it to the pool
 
         Returns:
-            Deferred[EmailPusher|HttpPusher]
+            EmailPusher|HttpPusher
         """
 
         time_now_msec = self.clock.time_msec()
@@ -124,9 +121,9 @@ class PusherPool:
         # create the pusher setting last_stream_ordering to the current maximum
         # stream ordering in event_push_actions, so it will process
         # pushes from this point onwards.
-        last_stream_ordering = yield self.store.get_latest_push_action_stream_ordering()
+        last_stream_ordering = await self.store.get_latest_push_action_stream_ordering()
 
-        yield self.store.add_pusher(
+        await self.store.add_pusher(
             user_id=user_id,
             access_token=access_token,
             kind=kind,
@@ -140,15 +137,14 @@ class PusherPool:
             last_stream_ordering=last_stream_ordering,
             profile_tag=profile_tag,
         )
-        pusher = yield self.start_pusher_by_id(app_id, pushkey, user_id)
+        pusher = await self.start_pusher_by_id(app_id, pushkey, user_id)
 
         return pusher
 
-    @defer.inlineCallbacks
-    def remove_pushers_by_app_id_and_pushkey_not_user(
+    async def remove_pushers_by_app_id_and_pushkey_not_user(
         self, app_id, pushkey, not_user_id
     ):
-        to_remove = yield self.store.get_pushers_by_app_id_and_pushkey(app_id, pushkey)
+        to_remove = await self.store.get_pushers_by_app_id_and_pushkey(app_id, pushkey)
         for p in to_remove:
             if p["user_name"] != not_user_id:
                 logger.info(
@@ -157,10 +153,9 @@ class PusherPool:
                     pushkey,
                     p["user_name"],
                 )
-                yield self.remove_pusher(p["app_id"], p["pushkey"], p["user_name"])
+                await self.remove_pusher(p["app_id"], p["pushkey"], p["user_name"])
 
-    @defer.inlineCallbacks
-    def remove_pushers_by_access_token(self, user_id, access_tokens):
+    async def remove_pushers_by_access_token(self, user_id, access_tokens):
         """Remove the pushers for a given user corresponding to a set of
         access_tokens.
 
@@ -173,7 +168,7 @@ class PusherPool:
             return
 
         tokens = set(access_tokens)
-        for p in (yield self.store.get_pushers_by_user_id(user_id)):
+        for p in await self.store.get_pushers_by_user_id(user_id):
             if p["access_token"] in tokens:
                 logger.info(
                     "Removing pusher for app id %s, pushkey %s, user %s",
@@ -181,16 +176,15 @@ class PusherPool:
                     p["pushkey"],
                     p["user_name"],
                 )
-                yield self.remove_pusher(p["app_id"], p["pushkey"], p["user_name"])
+                await self.remove_pusher(p["app_id"], p["pushkey"], p["user_name"])
 
-    @defer.inlineCallbacks
-    def on_new_notifications(self, min_stream_id, max_stream_id):
+    async def on_new_notifications(self, min_stream_id, max_stream_id):
         if not self.pushers:
             # nothing to do here.
             return
 
         try:
-            users_affected = yield self.store.get_push_action_users_in_range(
+            users_affected = await self.store.get_push_action_users_in_range(
                 min_stream_id, max_stream_id
             )
 
@@ -202,8 +196,7 @@ class PusherPool:
         except Exception:
             logger.exception("Exception in pusher on_new_notifications")
 
-    @defer.inlineCallbacks
-    def on_new_receipts(self, min_stream_id, max_stream_id, affected_room_ids):
+    async def on_new_receipts(self, min_stream_id, max_stream_id, affected_room_ids):
         if not self.pushers:
             # nothing to do here.
             return
@@ -211,7 +204,7 @@ class PusherPool:
         try:
             # Need to subtract 1 from the minimum because the lower bound here
             # is not inclusive
-            users_affected = yield self.store.get_users_sent_receipts_between(
+            users_affected = await self.store.get_users_sent_receipts_between(
                 min_stream_id - 1, max_stream_id
             )
 
@@ -223,12 +216,11 @@ class PusherPool:
         except Exception:
             logger.exception("Exception in pusher on_new_receipts")
 
-    @defer.inlineCallbacks
-    def start_pusher_by_id(self, app_id, pushkey, user_id):
+    async def start_pusher_by_id(self, app_id, pushkey, user_id):
         """Look up the details for the given pusher, and start it
 
         Returns:
-            Deferred[EmailPusher|HttpPusher|None]: The pusher started, if any
+            EmailPusher|HttpPusher|None: The pusher started, if any
         """
         if not self._should_start_pushers:
             return
@@ -236,7 +228,7 @@ class PusherPool:
         if not self._pusher_shard_config.should_handle(self._instance_name, user_id):
             return
 
-        resultlist = yield self.store.get_pushers_by_app_id_and_pushkey(app_id, pushkey)
+        resultlist = await self.store.get_pushers_by_app_id_and_pushkey(app_id, pushkey)
 
         pusher_dict = None
         for r in resultlist:
@@ -245,34 +237,29 @@ class PusherPool:
 
         pusher = None
         if pusher_dict:
-            pusher = yield self._start_pusher(pusher_dict)
+            pusher = await self._start_pusher(pusher_dict)
 
         return pusher
 
-    @defer.inlineCallbacks
-    def _start_pushers(self):
+    async def _start_pushers(self) -> None:
         """Start all the pushers
-
-        Returns:
-            Deferred
         """
-        pushers = yield self.store.get_all_pushers()
+        pushers = await self.store.get_all_pushers()
 
         # Stagger starting up the pushers so we don't completely drown the
         # process on start up.
-        yield concurrently_execute(self._start_pusher, pushers, 10)
+        await concurrently_execute(self._start_pusher, pushers, 10)
 
         logger.info("Started pushers")
 
-    @defer.inlineCallbacks
-    def _start_pusher(self, pusherdict):
+    async def _start_pusher(self, pusherdict):
         """Start the given pusher
 
         Args:
             pusherdict (dict): dict with the values pulled from the db table
 
         Returns:
-            Deferred[EmailPusher|HttpPusher]
+            EmailPusher|HttpPusher
         """
         if not self._pusher_shard_config.should_handle(
             self._instance_name, pusherdict["user_name"]
@@ -315,7 +302,7 @@ class PusherPool:
         user_id = pusherdict["user_name"]
         last_stream_ordering = pusherdict["last_stream_ordering"]
         if last_stream_ordering:
-            have_notifs = yield self.store.get_if_maybe_push_in_range_for_user(
+            have_notifs = await self.store.get_if_maybe_push_in_range_for_user(
                 user_id, last_stream_ordering
             )
         else:
@@ -327,8 +314,7 @@ class PusherPool:
 
         return p
 
-    @defer.inlineCallbacks
-    def remove_pusher(self, app_id, pushkey, user_id):
+    async def remove_pusher(self, app_id, pushkey, user_id):
         appid_pushkey = "%s:%s" % (app_id, pushkey)
 
         byuser = self.pushers.get(user_id, {})
@@ -340,6 +326,6 @@ class PusherPool:
 
             synapse_pushers.labels(type(pusher).__name__, pusher.app_id).dec()
 
-        yield self.store.delete_pusher_by_app_id_pushkey_user_id(
+        await self.store.delete_pusher_by_app_id_pushkey_user_id(
             app_id, pushkey, user_id
         )

synapse/python_dependencies.py

@@ -43,7 +43,7 @@ REQUIREMENTS = [
     "jsonschema>=2.5.1",
     "frozendict>=1",
     "unpaddedbase64>=1.1.0",
-    "canonicaljson>=1.1.3",
+    "canonicaljson>=1.2.0",
     # we use the type definitions added in signedjson 1.1.
     "signedjson>=1.1.0",
     "pynacl>=1.2.1",

synapse/replication/http/_base.py

@@ -20,8 +20,6 @@ import urllib
 from inspect import signature
 from typing import Dict, List, Tuple
 
-from twisted.internet import defer
-
 from synapse.api.errors import (
     CodeMessageException,
     HttpResponseException,
@@ -101,7 +99,7 @@ class ReplicationEndpoint(object):
         assert self.METHOD in ("PUT", "POST", "GET")
 
     @abc.abstractmethod
-    def _serialize_payload(**kwargs):
+    async def _serialize_payload(**kwargs):
         """Static method that is called when creating a request.
 
         Concrete implementations should have explicit parameters (rather than
@@ -110,9 +108,8 @@ class ReplicationEndpoint(object):
         argument list.
 
         Returns:
-            Deferred[dict]|dict: If POST/PUT request then dictionary must be
-            JSON serialisable, otherwise must be appropriate for adding as
-            query args.
+            dict: If POST/PUT request then dictionary must be JSON serialisable,
+            otherwise must be appropriate for adding as query args.
         """
         return {}
 
@@ -144,8 +141,7 @@ class ReplicationEndpoint(object):
         instance_map = hs.config.worker.instance_map
 
         @trace(opname="outgoing_replication_request")
-        @defer.inlineCallbacks
-        def send_request(instance_name="master", **kwargs):
+        async def send_request(instance_name="master", **kwargs):
             if instance_name == local_instance_name:
                 raise Exception("Trying to send HTTP request to self")
             if instance_name == "master":
@@ -159,7 +155,7 @@ class ReplicationEndpoint(object):
                     "Instance %r not in 'instance_map' config" % (instance_name,)
                 )
 
-            data = yield cls._serialize_payload(**kwargs)
+            data = await cls._serialize_payload(**kwargs)
 
             url_args = [
                 urllib.parse.quote(kwargs[name], safe="") for name in cls.PATH_ARGS
@@ -197,7 +193,7 @@ class ReplicationEndpoint(object):
                     headers = {}  # type: Dict[bytes, List[bytes]]
                     inject_active_span_byte_dict(headers, None, check_destination=False)
                     try:
-                        result = yield request_func(uri, data, headers=headers)
+                        result = await request_func(uri, data, headers=headers)
                         break
                     except CodeMessageException as e:
                         if e.code != 504 or not cls.RETRY_ON_TIMEOUT:
@@ -207,7 +203,7 @@ class ReplicationEndpoint(object):
 
                     # If we timed out we probably don't need to worry about backing
                     # off too much, but lets just wait a little anyway.
-                    yield clock.sleep(1)
+                    await clock.sleep(1)
             except HttpResponseException as e:
                 # We convert to SynapseError as we know that it was a SynapseError
                 # on the master process that we should send to the client. (And

synapse/replication/http/devices.py

@@ -60,7 +60,7 @@ class ReplicationUserDevicesResyncRestServlet(ReplicationEndpoint):
         self.clock = hs.get_clock()
 
     @staticmethod
-    def _serialize_payload(user_id):
+    async def _serialize_payload(user_id):
         return {}
 
     async def _handle_request(self, request, user_id):

synapse/replication/http/federation.py

@@ -15,8 +15,6 @@
 
 import logging
 
-from twisted.internet import defer
-
 from synapse.api.room_versions import KNOWN_ROOM_VERSIONS
 from synapse.events import make_event_from_dict
 from synapse.events.snapshot import EventContext
@@ -67,8 +65,7 @@ class ReplicationFederationSendEventsRestServlet(ReplicationEndpoint):
         self.federation_handler = hs.get_handlers().federation_handler
 
     @staticmethod
-    @defer.inlineCallbacks
-    def _serialize_payload(store, event_and_contexts, backfilled):
+    async def _serialize_payload(store, event_and_contexts, backfilled):
         """
         Args:
             store
@@ -78,7 +75,7 @@ class ReplicationFederationSendEventsRestServlet(ReplicationEndpoint):
         """
         event_payloads = []
         for event, context in event_and_contexts:
-            serialized_context = yield context.serialize(event, store)
+            serialized_context = await context.serialize(event, store)
 
             event_payloads.append(
                 {
@@ -154,7 +151,7 @@ class ReplicationFederationSendEduRestServlet(ReplicationEndpoint):
         self.registry = hs.get_federation_registry()
 
     @staticmethod
-    def _serialize_payload(edu_type, origin, content):
+    async def _serialize_payload(edu_type, origin, content):
         return {"origin": origin, "content": content}
 
     async def _handle_request(self, request, edu_type):
@@ -197,7 +194,7 @@ class ReplicationGetQueryRestServlet(ReplicationEndpoint):
         self.registry = hs.get_federation_registry()
 
     @staticmethod
-    def _serialize_payload(query_type, args):
+    async def _serialize_payload(query_type, args):
         """
         Args:
             query_type (str)
@@ -238,7 +235,7 @@ class ReplicationCleanRoomRestServlet(ReplicationEndpoint):
         self.store = hs.get_datastore()
 
     @staticmethod
-    def _serialize_payload(room_id, args):
+    async def _serialize_payload(room_id, args):
         """
         Args:
             room_id (str)
@@ -273,7 +270,7 @@ class ReplicationStoreRoomOnInviteRestServlet(ReplicationEndpoint):
         self.store = hs.get_datastore()
 
     @staticmethod
-    def _serialize_payload(room_id, room_version):
+    async def _serialize_payload(room_id, room_version):
         return {"room_version": room_version.identifier}
 
     async def _handle_request(self, request, room_id):

synapse/replication/http/login.py

@@ -36,7 +36,7 @@ class RegisterDeviceReplicationServlet(ReplicationEndpoint):
         self.registration_handler = hs.get_registration_handler()
 
     @staticmethod
-    def _serialize_payload(user_id, device_id, initial_display_name, is_guest):
+    async def _serialize_payload(user_id, device_id, initial_display_name, is_guest):
         """
         Args:
             device_id (str|None): Device ID to use, if None a new one is

synapse/replication/http/membership.py

@@ -52,7 +52,9 @@ class ReplicationRemoteJoinRestServlet(ReplicationEndpoint):
         self.clock = hs.get_clock()
 
     @staticmethod
-    def _serialize_payload(requester, room_id, user_id, remote_room_hosts, content):
+    async def _serialize_payload(
+        requester, room_id, user_id, remote_room_hosts, content
+    ):
         """
         Args:
             requester(Requester)
@@ -112,7 +114,7 @@ class ReplicationRemoteRejectInviteRestServlet(ReplicationEndpoint):
         self.member_handler = hs.get_room_member_handler()
 
     @staticmethod
-    def _serialize_payload(  # type: ignore
+    async def _serialize_payload(  # type: ignore
         invite_event_id: str,
         txn_id: Optional[str],
         requester: Requester,
@@ -174,7 +176,7 @@ class ReplicationUserJoinedLeftRoomRestServlet(ReplicationEndpoint):
         self.distributor = hs.get_distributor()
 
     @staticmethod
-    def _serialize_payload(room_id, user_id, change):
+    async def _serialize_payload(room_id, user_id, change):
         """
         Args:
             room_id (str)

synapse/replication/http/presence.py

@@ -50,7 +50,7 @@ class ReplicationBumpPresenceActiveTime(ReplicationEndpoint):
         self._presence_handler = hs.get_presence_handler()
 
     @staticmethod
-    def _serialize_payload(user_id):
+    async def _serialize_payload(user_id):
         return {}
 
     async def _handle_request(self, request, user_id):
@@ -92,7 +92,7 @@ class ReplicationPresenceSetState(ReplicationEndpoint):
         self._presence_handler = hs.get_presence_handler()
 
     @staticmethod
-    def _serialize_payload(user_id, state, ignore_status_msg=False):
+    async def _serialize_payload(user_id, state, ignore_status_msg=False):
         return {
             "state": state,
             "ignore_status_msg": ignore_status_msg,

synapse/replication/http/register.py

@@ -34,7 +34,7 @@ class ReplicationRegisterServlet(ReplicationEndpoint):
         self.registration_handler = hs.get_registration_handler()
 
     @staticmethod
-    def _serialize_payload(
+    async def _serialize_payload(
         user_id,
         password_hash,
         was_guest,
@@ -105,7 +105,7 @@ class ReplicationPostRegisterActionsServlet(ReplicationEndpoint):
         self.registration_handler = hs.get_registration_handler()
 
     @staticmethod
-    def _serialize_payload(user_id, auth_result, access_token):
+    async def _serialize_payload(user_id, auth_result, access_token):
         """
         Args:
             user_id (str): The user ID that consented

synapse/replication/http/send_event.py

@@ -15,8 +15,6 @@
 
 import logging
 
-from twisted.internet import defer
-
 from synapse.api.room_versions import KNOWN_ROOM_VERSIONS
 from synapse.events import make_event_from_dict
 from synapse.events.snapshot import EventContext
@@ -62,8 +60,7 @@ class ReplicationSendEventRestServlet(ReplicationEndpoint):
         self.clock = hs.get_clock()
 
     @staticmethod
-    @defer.inlineCallbacks
-    def _serialize_payload(
+    async def _serialize_payload(
         event_id, store, event, context, requester, ratelimit, extra_users
     ):
         """
@@ -77,7 +74,7 @@ class ReplicationSendEventRestServlet(ReplicationEndpoint):
             extra_users (list(UserID)): Any extra users to notify about event
         """
 
-        serialized_context = yield context.serialize(event, store)
+        serialized_context = await context.serialize(event, store)
 
         payload = {
             "event": event.get_pdu_json(),

synapse/replication/http/streams.py

@@ -54,7 +54,7 @@ class ReplicationGetStreamUpdates(ReplicationEndpoint):
         self.streams = hs.get_replication_streams()
 
     @staticmethod
-    def _serialize_payload(stream_name, from_token, upto_token):
+    async def _serialize_payload(stream_name, from_token, upto_token):
         return {"from_token": from_token, "upto_token": upto_token}
 
     async def _handle_request(self, request, stream_name):

synapse/rest/admin/rooms.py

@@ -103,6 +103,14 @@ class DeleteRoomRestServlet(RestServlet):
                 Codes.BAD_JSON,
             )
 
+        purge = content.get("purge", True)
+        if not isinstance(purge, bool):
+            raise SynapseError(
+                HTTPStatus.BAD_REQUEST,
+                "Param 'purge' must be a boolean, if given",
+                Codes.BAD_JSON,
+            )
+
         ret = await self.room_shutdown_handler.shutdown_room(
             room_id=room_id,
             new_room_user_id=content.get("new_room_user_id"),
@@ -113,7 +121,8 @@ class DeleteRoomRestServlet(RestServlet):
         )
 
         # Purge room
-        await self.pagination_handler.purge_room(room_id)
+        if purge:
+            await self.pagination_handler.purge_room(room_id)
 
         return (200, ret)
 

synapse/rest/client/v1/room.py

@@ -444,7 +444,7 @@ class RoomMemberListRestServlet(RestServlet):
 
     async def on_GET(self, request, room_id):
         # TODO support Pagination stream API (limit/tokens)
-        requester = await self.auth.get_user_by_req(request)
+        requester = await self.auth.get_user_by_req(request, allow_guest=True)
         handler = self.message_handler
 
         # request the state as of a given event, as identified by a stream token,

synapse/rest/client/v2_alpha/sync.py

@@ -426,6 +426,7 @@ class SyncRestServlet(RestServlet):
             result["ephemeral"] = {"events": ephemeral_events}
             result["unread_notifications"] = room.unread_notifications
             result["summary"] = room.summary
+            result["org.matrix.msc2654.unread_count"] = room.unread_count
 
         return result