Prepare for authenticated media freeze (#17433)

As part of the rollout of
[MSC3916](https://github.com/matrix-org/matrix-spec-proposals/blob/main/proposals/3916-authentication-for-media.md)
this PR adds support for designating authenticated media and ensuring
that authenticated media is not served over unauthenticated endpoints.

9 files changed, 140 insertions, 12 deletions

diff --git a/synapse/_scripts/synapse_port_db.py b/synapse/_scripts/synapse_port_db.py
index 3bb4a34938..5c6db8118f 100755
--- a/synapse/_scripts/synapse_port_db.py
+++ b/synapse/_scripts/synapse_port_db.py
@@ -119,18 +119,19 @@ BOOLEAN_COLUMNS = {
     "e2e_room_keys": ["is_verified"],
     "event_edges": ["is_state"],
     "events": ["processed", "outlier", "contains_url"],
-    "local_media_repository": ["safe_from_quarantine"],
+    "local_media_repository": ["safe_from_quarantine", "authenticated"],
+    "per_user_experimental_features": ["enabled"],
     "presence_list": ["accepted"],
     "presence_stream": ["currently_active"],
     "public_room_list_stream": ["visibility"],
     "pushers": ["enabled"],
     "redactions": ["have_censored"],
+    "remote_media_cache": ["authenticated"],
     "room_stats_state": ["is_federatable"],
     "rooms": ["is_public", "has_auth_chain_index"],
     "users": ["shadow_banned", "approved", "locked", "suspended"],
     "un_partial_stated_event_stream": ["rejection_status_changed"],
     "users_who_share_rooms": ["share_private"],
-    "per_user_experimental_features": ["enabled"],
 }
 
 
diff --git a/synapse/config/repository.py b/synapse/config/repository.py
index dc0e93ffa1..97ce6de528 100644
--- a/synapse/config/repository.py
+++ b/synapse/config/repository.py
@@ -272,6 +272,10 @@ class ContentRepositoryConfig(Config):
                 remote_media_lifetime
             )
 
+        self.enable_authenticated_media = config.get(
+            "enable_authenticated_media", False
+        )
+
     def generate_config_section(self, data_dir_path: str, **kwargs: Any) -> str:
         assert data_dir_path is not None
         media_store = os.path.join(data_dir_path, "media_store")
diff --git a/synapse/media/media_repository.py b/synapse/media/media_repository.py
index 87c929eb20..8bc92305fe 100644
--- a/synapse/media/media_repository.py
+++ b/synapse/media/media_repository.py
@@ -430,6 +430,7 @@ class MediaRepository:
         media_id: str,
         name: Optional[str],
         max_timeout_ms: int,
+        allow_authenticated: bool = True,
         federation: bool = False,
     ) -> None:
         """Responds to requests for local media, if exists, or returns 404.
@@ -442,6 +443,7 @@ class MediaRepository:
                 the filename in the Content-Disposition header of the response.
             max_timeout_ms: the maximum number of milliseconds to wait for the
                 media to be uploaded.
+            allow_authenticated: whether media marked as authenticated may be served to this request
             federation: whether the local media being fetched is for a federation request
 
         Returns:
@@ -451,6 +453,10 @@ class MediaRepository:
         if not media_info:
             return
 
+        if self.hs.config.media.enable_authenticated_media and not allow_authenticated:
+            if media_info.authenticated:
+                raise NotFoundError()
+
         self.mark_recently_accessed(None, media_id)
 
         media_type = media_info.media_type
@@ -481,6 +487,7 @@ class MediaRepository:
         max_timeout_ms: int,
         ip_address: str,
         use_federation_endpoint: bool,
+        allow_authenticated: bool = True,
     ) -> None:
         """Respond to requests for remote media.
 
@@ -495,6 +502,8 @@ class MediaRepository:
             ip_address: the IP address of the requester
             use_federation_endpoint: whether to request the remote media over the new
                 federation `/download` endpoint
+            allow_authenticated: whether media marked as authenticated may be served to this
+                request
 
         Returns:
             Resolves once a response has successfully been written to request
@@ -526,6 +535,7 @@ class MediaRepository:
                 self.download_ratelimiter,
                 ip_address,
                 use_federation_endpoint,
+                allow_authenticated,
             )
 
         # We deliberately stream the file outside the lock
@@ -548,6 +558,7 @@ class MediaRepository:
         max_timeout_ms: int,
         ip_address: str,
         use_federation: bool,
+        allow_authenticated: bool,
     ) -> RemoteMedia:
         """Gets the media info associated with the remote file, downloading
         if necessary.
@@ -560,6 +571,8 @@ class MediaRepository:
             ip_address: IP address of the requester
             use_federation: if a download is necessary, whether to request the remote file
                 over the federation `/download` endpoint
+            allow_authenticated: whether media marked as authenticated may be served to this
+                request
 
         Returns:
             The media info of the file
@@ -581,6 +594,7 @@ class MediaRepository:
                 self.download_ratelimiter,
                 ip_address,
                 use_federation,
+                allow_authenticated,
             )
 
         # Ensure we actually use the responder so that it releases resources
@@ -598,6 +612,7 @@ class MediaRepository:
         download_ratelimiter: Ratelimiter,
         ip_address: str,
         use_federation_endpoint: bool,
+        allow_authenticated: bool,
     ) -> Tuple[Optional[Responder], RemoteMedia]:
         """Looks for media in local cache, if not there then attempt to
         download from remote server.
@@ -619,6 +634,11 @@ class MediaRepository:
         """
         media_info = await self.store.get_cached_remote_media(server_name, media_id)
 
+        if self.hs.config.media.enable_authenticated_media and not allow_authenticated:
+            # if it isn't cached then don't fetch it or if it's authenticated then don't serve it
+            if not media_info or media_info.authenticated:
+                raise NotFoundError()
+
         # file_id is the ID we use to track the file locally. If we've already
         # seen the file then reuse the existing ID, otherwise generate a new
         # one.
@@ -792,6 +812,11 @@ class MediaRepository:
 
         logger.info("Stored remote media in file %r", fname)
 
+        if self.hs.config.media.enable_authenticated_media:
+            authenticated = True
+        else:
+            authenticated = False
+
         return RemoteMedia(
             media_origin=server_name,
             media_id=media_id,
@@ -802,6 +827,7 @@ class MediaRepository:
             filesystem_id=file_id,
             last_access_ts=time_now_ms,
             quarantined_by=None,
+            authenticated=authenticated,
         )
 
     async def _federation_download_remote_file(
@@ -915,6 +941,11 @@ class MediaRepository:
 
         logger.debug("Stored remote media in file %r", fname)
 
+        if self.hs.config.media.enable_authenticated_media:
+            authenticated = True
+        else:
+            authenticated = False
+
         return RemoteMedia(
             media_origin=server_name,
             media_id=media_id,
@@ -925,6 +956,7 @@ class MediaRepository:
             filesystem_id=file_id,
             last_access_ts=time_now_ms,
             quarantined_by=None,
+            authenticated=authenticated,
         )
 
     def _get_thumbnail_requirements(
@@ -1030,7 +1062,12 @@ class MediaRepository:
             t_len = os.path.getsize(output_path)
 
             await self.store.store_local_thumbnail(
-                media_id, t_width, t_height, t_type, t_method, t_len
+                media_id,
+                t_width,
+                t_height,
+                t_type,
+                t_method,
+                t_len,
             )
 
             return output_path
diff --git a/synapse/media/thumbnailer.py b/synapse/media/thumbnailer.py
index 413a720e40..ef6aa8ccf5 100644
--- a/synapse/media/thumbnailer.py
+++ b/synapse/media/thumbnailer.py
@@ -26,7 +26,7 @@ from typing import TYPE_CHECKING, List, Optional, Tuple, Type
 
 from PIL import Image
 
-from synapse.api.errors import Codes, SynapseError, cs_error
+from synapse.api.errors import Codes, NotFoundError, SynapseError, cs_error
 from synapse.config.repository import THUMBNAIL_SUPPORTED_MEDIA_FORMAT_MAP
 from synapse.http.server import respond_with_json
 from synapse.http.site import SynapseRequest
@@ -274,6 +274,7 @@ class ThumbnailProvider:
         m_type: str,
         max_timeout_ms: int,
         for_federation: bool,
+        allow_authenticated: bool = True,
     ) -> None:
         media_info = await self.media_repo.get_local_media_info(
             request, media_id, max_timeout_ms
@@ -281,6 +282,12 @@ class ThumbnailProvider:
         if not media_info:
             return
 
+        # if the media the thumbnail is generated from is authenticated, don't serve the
+        # thumbnail over an unauthenticated endpoint
+        if self.hs.config.media.enable_authenticated_media and not allow_authenticated:
+            if media_info.authenticated:
+                raise NotFoundError()
+
         thumbnail_infos = await self.store.get_local_media_thumbnails(media_id)
         await self._select_and_respond_with_thumbnail(
             request,
@@ -307,14 +314,20 @@ class ThumbnailProvider:
         desired_type: str,
         max_timeout_ms: int,
         for_federation: bool,
+        allow_authenticated: bool = True,
     ) -> None:
         media_info = await self.media_repo.get_local_media_info(
             request, media_id, max_timeout_ms
         )
-
         if not media_info:
             return
 
+        # if the media the thumbnail is generated from is authenticated, don't serve the
+        # thumbnail over an unauthenticated endpoint
+        if self.hs.config.media.enable_authenticated_media and not allow_authenticated:
+            if media_info.authenticated:
+                raise NotFoundError()
+
         thumbnail_infos = await self.store.get_local_media_thumbnails(media_id)
         for info in thumbnail_infos:
             t_w = info.width == desired_width
@@ -381,14 +394,27 @@ class ThumbnailProvider:
         max_timeout_ms: int,
         ip_address: str,
         use_federation: bool,
+        allow_authenticated: bool = True,
     ) -> None:
         media_info = await self.media_repo.get_remote_media_info(
-            server_name, media_id, max_timeout_ms, ip_address, use_federation
+            server_name,
+            media_id,
+            max_timeout_ms,
+            ip_address,
+            use_federation,
+            allow_authenticated,
         )
         if not media_info:
             respond_404(request)
             return
 
+        # if the media the thumbnail is generated from is authenticated, don't serve the
+        # thumbnail over an unauthenticated endpoint
+        if self.hs.config.media.enable_authenticated_media and not allow_authenticated:
+            if media_info.authenticated:
+                respond_404(request)
+                return
+
         thumbnail_infos = await self.store.get_remote_media_thumbnails(
             server_name, media_id
         )
@@ -446,16 +472,28 @@ class ThumbnailProvider:
         max_timeout_ms: int,
         ip_address: str,
         use_federation: bool,
+        allow_authenticated: bool = True,
     ) -> None:
         # TODO: Don't download the whole remote file
         # We should proxy the thumbnail from the remote server instead of
         # downloading the remote file and generating our own thumbnails.
         media_info = await self.media_repo.get_remote_media_info(
-            server_name, media_id, max_timeout_ms, ip_address, use_federation
+            server_name,
+            media_id,
+            max_timeout_ms,
+            ip_address,
+            use_federation,
+            allow_authenticated,
         )
         if not media_info:
             return
 
+        # if the media the thumbnail is generated from is authenticated, don't serve the
+        # thumbnail over an unauthenticated endpoint
+        if self.hs.config.media.enable_authenticated_media and not allow_authenticated:
+            if media_info.authenticated:
+                raise NotFoundError()
+
         thumbnail_infos = await self.store.get_remote_media_thumbnails(
             server_name, media_id
         )
@@ -485,8 +523,8 @@ class ThumbnailProvider:
         file_id: str,
         url_cache: bool,
         for_federation: bool,
-        server_name: Optional[str] = None,
         media_info: Optional[LocalMedia] = None,
+        server_name: Optional[str] = None,
     ) -> None:
         """
         Respond to a request with an appropriate thumbnail from the previously generated thumbnails.
diff --git a/synapse/rest/media/download_resource.py b/synapse/rest/media/download_resource.py
index c32c626905..3c3f703667 100644
--- a/synapse/rest/media/download_resource.py
+++ b/synapse/rest/media/download_resource.py
@@ -84,7 +84,7 @@ class DownloadResource(RestServlet):
 
         if self._is_mine_server_name(server_name):
             await self.media_repo.get_local_media(
-                request, media_id, file_name, max_timeout_ms
+                request, media_id, file_name, max_timeout_ms, allow_authenticated=False
             )
         else:
             allow_remote = parse_boolean(request, "allow_remote", default=True)
@@ -106,4 +106,5 @@ class DownloadResource(RestServlet):
                 max_timeout_ms,
                 ip_address,
                 False,
+                allow_authenticated=False,
             )
diff --git a/synapse/rest/media/thumbnail_resource.py b/synapse/rest/media/thumbnail_resource.py
index 70354aa439..536fea4c32 100644
--- a/synapse/rest/media/thumbnail_resource.py
+++ b/synapse/rest/media/thumbnail_resource.py
@@ -96,6 +96,7 @@ class ThumbnailResource(RestServlet):
                     m_type,
                     max_timeout_ms,
                     False,
+                    allow_authenticated=False,
                 )
             else:
                 await self.thumbnail_provider.respond_local_thumbnail(
@@ -107,6 +108,7 @@ class ThumbnailResource(RestServlet):
                     m_type,
                     max_timeout_ms,
                     False,
+                    allow_authenticated=False,
                 )
             self.media_repo.mark_recently_accessed(None, media_id)
         else:
@@ -134,6 +136,7 @@ class ThumbnailResource(RestServlet):
                 m_type,
                 max_timeout_ms,
                 ip_address,
-                False,
+                use_federation=False,
+                allow_authenticated=False,
             )
             self.media_repo.mark_recently_accessed(server_name, media_id)
diff --git a/synapse/storage/databases/main/media_repository.py b/synapse/storage/databases/main/media_repository.py
index 6128332af8..7617fd3ad4 100644
--- a/synapse/storage/databases/main/media_repository.py
+++ b/synapse/storage/databases/main/media_repository.py
@@ -64,6 +64,7 @@ class LocalMedia:
     quarantined_by: Optional[str]
     safe_from_quarantine: bool
     user_id: Optional[str]
+    authenticated: Optional[bool]
 
 
 @attr.s(slots=True, frozen=True, auto_attribs=True)
@@ -77,6 +78,7 @@ class RemoteMedia:
     created_ts: int
     last_access_ts: int
     quarantined_by: Optional[str]
+    authenticated: Optional[bool]
 
 
 @attr.s(slots=True, frozen=True, auto_attribs=True)
@@ -218,6 +220,7 @@ class MediaRepositoryStore(MediaRepositoryBackgroundUpdateStore):
                 "last_access_ts",
                 "safe_from_quarantine",
                 "user_id",
+                "authenticated",
             ),
             allow_none=True,
             desc="get_local_media",
@@ -235,6 +238,7 @@ class MediaRepositoryStore(MediaRepositoryBackgroundUpdateStore):
             last_access_ts=row[6],
             safe_from_quarantine=row[7],
             user_id=row[8],
+            authenticated=row[9],
         )
 
     async def get_local_media_by_user_paginate(
@@ -290,7 +294,8 @@ class MediaRepositoryStore(MediaRepositoryBackgroundUpdateStore):
                     last_access_ts,
                     quarantined_by,
                     safe_from_quarantine,
-                    user_id
+                    user_id,
+                    authenticated
                 FROM local_media_repository
                 WHERE user_id = ?
                 ORDER BY {order_by_column} {order}, media_id ASC
@@ -314,6 +319,7 @@ class MediaRepositoryStore(MediaRepositoryBackgroundUpdateStore):
                     quarantined_by=row[7],
                     safe_from_quarantine=bool(row[8]),
                     user_id=row[9],
+                    authenticated=row[10],
                 )
                 for row in txn
             ]
@@ -417,12 +423,18 @@ class MediaRepositoryStore(MediaRepositoryBackgroundUpdateStore):
         time_now_ms: int,
         user_id: UserID,
     ) -> None:
+        if self.hs.config.media.enable_authenticated_media:
+            authenticated = True
+        else:
+            authenticated = False
+
         await self.db_pool.simple_insert(
             "local_media_repository",
             {
                 "media_id": media_id,
                 "created_ts": time_now_ms,
                 "user_id": user_id.to_string(),
+                "authenticated": authenticated,
             },
             desc="store_local_media_id",
         )
@@ -438,6 +450,11 @@ class MediaRepositoryStore(MediaRepositoryBackgroundUpdateStore):
         user_id: UserID,
         url_cache: Optional[str] = None,
     ) -> None:
+        if self.hs.config.media.enable_authenticated_media:
+            authenticated = True
+        else:
+            authenticated = False
+
         await self.db_pool.simple_insert(
             "local_media_repository",
             {
@@ -448,6 +465,7 @@ class MediaRepositoryStore(MediaRepositoryBackgroundUpdateStore):
                 "media_length": media_length,
                 "user_id": user_id.to_string(),
                 "url_cache": url_cache,
+                "authenticated": authenticated,
             },
             desc="store_local_media",
         )
@@ -638,6 +656,7 @@ class MediaRepositoryStore(MediaRepositoryBackgroundUpdateStore):
                 "filesystem_id",
                 "last_access_ts",
                 "quarantined_by",
+                "authenticated",
             ),
             allow_none=True,
             desc="get_cached_remote_media",
@@ -654,6 +673,7 @@ class MediaRepositoryStore(MediaRepositoryBackgroundUpdateStore):
             filesystem_id=row[4],
             last_access_ts=row[5],
             quarantined_by=row[6],
+            authenticated=row[7],
         )
 
     async def store_cached_remote_media(
@@ -666,6 +686,11 @@ class MediaRepositoryStore(MediaRepositoryBackgroundUpdateStore):
         upload_name: Optional[str],
         filesystem_id: str,
     ) -> None:
+        if self.hs.config.media.enable_authenticated_media:
+            authenticated = True
+        else:
+            authenticated = False
+
         await self.db_pool.simple_insert(
             "remote_media_cache",
             {
@@ -677,6 +702,7 @@ class MediaRepositoryStore(MediaRepositoryBackgroundUpdateStore):
                 "upload_name": upload_name,
                 "filesystem_id": filesystem_id,
                 "last_access_ts": time_now_ms,
+                "authenticated": authenticated,
             },
             desc="store_cached_remote_media",
         )
diff --git a/synapse/storage/schema/__init__.py b/synapse/storage/schema/__init__.py
index 0dc5d24249..581d00346b 100644
--- a/synapse/storage/schema/__init__.py
+++ b/synapse/storage/schema/__init__.py
@@ -19,7 +19,7 @@
 #
 #
 
-SCHEMA_VERSION = 85  # remember to update the list below when updating
+SCHEMA_VERSION = 86  # remember to update the list below when updating
 """Represents the expectations made by the codebase about the database schema
 
 This should be incremented whenever the codebase changes its requirements on the
@@ -139,6 +139,9 @@ Changes in SCHEMA_VERSION = 84
 
 Changes in SCHEMA_VERSION = 85
     - Add a column `suspended` to the `users` table
+
+Changes in SCHEMA_VERSION = 86
+    - Add a column `authenticated` to the tables `local_media_repository` and `remote_media_cache`
 """
 
 
diff --git a/synapse/storage/schema/main/delta/86/01_authenticate_media.sql b/synapse/storage/schema/main/delta/86/01_authenticate_media.sql
new file mode 100644
index 0000000000..c1ac01ae95
--- /dev/null
+++ b/synapse/storage/schema/main/delta/86/01_authenticate_media.sql
@@ -0,0 +1,15 @@
+--
+-- This file is licensed under the Affero General Public License (AGPL) version 3.
+--
+-- Copyright (C) 2024 New Vector, Ltd
+--
+-- This program is free software: you can redistribute it and/or modify
+-- it under the terms of the GNU Affero General Public License as
+-- published by the Free Software Foundation, either version 3 of the
+-- License, or (at your option) any later version.
+--
+-- See the GNU Affero General Public License for more details:
+-- <https://www.gnu.org/licenses/agpl-3.0.html>.
+
+ALTER TABLE remote_media_cache ADD COLUMN authenticated BOOLEAN DEFAULT FALSE NOT NULL;
+ALTER TABLE local_media_repository ADD COLUMN authenticated BOOLEAN DEFAULT FALSE NOT NULL;