From 84cf3e47a0318aba51d9f830d5e724182c5d93c4 Mon Sep 17 00:00:00 2001
From: Erik Johnston <erik@matrix.org>
Date: Fri, 28 May 2021 16:28:01 +0100
Subject: [PATCH] Allow response of `/send_join` to be larger. (#10093)

Fixes #10087.
---
 changelog.d/10093.bugfix               |  1 +
 synapse/federation/transport/client.py |  7 +++++++
 synapse/http/matrixfederationclient.py | 14 +++++++++++++-
 3 files changed, 21 insertions(+), 1 deletion(-)
 create mode 100644 changelog.d/10093.bugfix

diff --git a/changelog.d/10093.bugfix b/changelog.d/10093.bugfix
new file mode 100644
index 0000000000..e50de4b2ea
--- /dev/null
+++ b/changelog.d/10093.bugfix
@@ -0,0 +1 @@
+Fix HTTP response size limit to allow joining very large rooms over federation.
diff --git a/synapse/federation/transport/client.py b/synapse/federation/transport/client.py
index e93ab83f7f..5b4f5d17f7 100644
--- a/synapse/federation/transport/client.py
+++ b/synapse/federation/transport/client.py
@@ -35,6 +35,11 @@ from synapse.types import JsonDict
 
 logger = logging.getLogger(__name__)
 
+# Send join responses can be huge, so we set a separate limit here. The response
+# is parsed in a streaming manner, which helps alleviate the issue of memory
+# usage a bit.
+MAX_RESPONSE_SIZE_SEND_JOIN = 500 * 1024 * 1024
+
 
 class TransportLayerClient:
     """Sends federation HTTP requests to other servers"""
@@ -261,6 +266,7 @@ class TransportLayerClient:
             path=path,
             data=content,
             parser=SendJoinParser(room_version, v1_api=True),
+            max_response_size=MAX_RESPONSE_SIZE_SEND_JOIN,
         )
 
         return response
@@ -276,6 +282,7 @@ class TransportLayerClient:
             path=path,
             data=content,
             parser=SendJoinParser(room_version, v1_api=False),
+            max_response_size=MAX_RESPONSE_SIZE_SEND_JOIN,
         )
 
         return response
diff --git a/synapse/http/matrixfederationclient.py b/synapse/http/matrixfederationclient.py
index f5503b394b..1998990a14 100644
--- a/synapse/http/matrixfederationclient.py
+++ b/synapse/http/matrixfederationclient.py
@@ -205,6 +205,7 @@ async def _handle_response(
     response: IResponse,
     start_ms: int,
     parser: ByteParser[T],
+    max_response_size: Optional[int] = None,
 ) -> T:
     """
     Reads the body of a response with a timeout and sends it to a parser
@@ -216,15 +217,20 @@ async def _handle_response(
         response: response to the request
         start_ms: Timestamp when request was made
         parser: The parser for the response
+        max_response_size: The maximum size to read from the response, if None
+            uses the default.
 
     Returns:
         The parsed response
     """
 
+    if max_response_size is None:
+        max_response_size = MAX_RESPONSE_SIZE
+
     try:
         check_content_type_is(response.headers, parser.CONTENT_TYPE)
 
-        d = read_body_with_max_size(response, parser, MAX_RESPONSE_SIZE)
+        d = read_body_with_max_size(response, parser, max_response_size)
         d = timeout_deferred(d, timeout=timeout_sec, reactor=reactor)
 
         length = await make_deferred_yieldable(d)
@@ -735,6 +741,7 @@ class MatrixFederationHttpClient:
         backoff_on_404: bool = False,
         try_trailing_slash_on_400: bool = False,
         parser: Literal[None] = None,
+        max_response_size: Optional[int] = None,
     ) -> Union[JsonDict, list]:
         ...
 
@@ -752,6 +759,7 @@ class MatrixFederationHttpClient:
         backoff_on_404: bool = False,
         try_trailing_slash_on_400: bool = False,
         parser: Optional[ByteParser[T]] = None,
+        max_response_size: Optional[int] = None,
     ) -> T:
         ...
 
@@ -768,6 +776,7 @@ class MatrixFederationHttpClient:
         backoff_on_404: bool = False,
         try_trailing_slash_on_400: bool = False,
         parser: Optional[ByteParser] = None,
+        max_response_size: Optional[int] = None,
     ):
         """Sends the specified json data using PUT
 
@@ -803,6 +812,8 @@ class MatrixFederationHttpClient:
                 enabled.
             parser: The parser to use to decode the response. Defaults to
                 parsing as JSON.
+            max_response_size: The maximum size to read from the response, if None
+                uses the default.
 
         Returns:
             Succeeds when we get a 2xx HTTP response. The
@@ -853,6 +864,7 @@ class MatrixFederationHttpClient:
             response,
             start_ms,
             parser=parser,
+            max_response_size=max_response_size,
         )
 
         return body