feat: Encrypted state events (MSC3414) #5456

kaylendog · 2025-07-28T15:06:46Z

Implements support for encrypted state events under m.room.encrypted, as outlined in MSC3414.

(base) Modifies state_events::sync::dispatch to require E2EE when end-to-end encryption is enabled, using it to decrypt m.room.encrypted state events when received.
(base) Declares m.room.encrypted as required state by default.
(crypto) Introduces encrypt_state_event and encrypt_state_event_raw to OlmMachine and OutboundGroupSession;
(sdk) Modifies Room::send_state_event and Room::send_state_event_Raw to return SendStateEvent and SendStateEventRaw futures respectifely, that each function similarly to their message-like counterparts.
(integration) Add an integration test for encrypted room renaming.

richvdh · 2025-07-28T16:23:32Z

This is part of #5397

crates/matrix-sdk-base/src/room/encryption.rs

codspeed-hq · 2025-07-30T13:41:57Z

CodSpeed Performance Report

Merging #5456 will not alter performance

_{Comparing kaylendog:kaylendog/encrypted-state-events-compat (61d40bd) with main (669ebf2)}

Summary

✅ 31 untouched benchmarks

zecakeh · 2025-08-01T10:18:18Z

For information, #5473 upgrades Ruma and brings in the encrypted state events changes.

Cargo.toml

codecov · 2025-08-04T16:55:06Z

Codecov Report

❌ Patch coverage is 80.78818% with 39 lines in your changes missing coverage. Please review.
✅ Project coverage is 88.56%. Comparing base (669ebf2) to head (a8b7d11).
⚠️ Report is 16 commits behind head on main.
✅ All tests successful. No failed tests found.

Files with missing lines	Patch %	Lines
crates/matrix-sdk/src/room/futures.rs	79.43%	13 Missing and 9 partials ⚠️
crates/matrix-sdk-crypto/src/machine/mod.rs	44.44%	15 Missing ⚠️
...k-crypto/src/session_manager/group_sessions/mod.rs	87.50%	0 Missing and 2 partials ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #5456      +/-   ##
==========================================
- Coverage   88.57%   88.56%   -0.01%     
==========================================
  Files         339      339              
  Lines       93603    93800     +197     
  Branches    93603    93800     +197     
==========================================
+ Hits        82910    83078     +168     
- Misses       6560     6577      +17     
- Partials     4133     4145      +12

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

labs/multiverse/src/widgets/room_view/mod.rs

testing/matrix-sdk-integration-testing/src/tests/e2ee/state_events.rs

Signed-off-by: Skye Elliot <[email protected]>

…macro Signed-off-by: Skye Elliot <[email protected]>

Signed-off-by: Skye Elliot <[email protected]>

richvdh

General comments:

Yay! excites! It generally looks very sensible.
I wonder if all this functionality should be behind compile-time cfg guards, since it's entirely unspecced. @poljar: any thoughts on this?
For future reference: your commit comments could do with a little more detail. Try and explain how the bit of functionality that bit introduces fits in with the whole.
Before this is ready to land in the rust SDK, I think we need some more (or, indeed, any) unit tests, which exercise individual parts of the functionality without rolling it into a big integration test.

For sure an integration test is nice to make sure it all works as a whole, but integration tests are slooow and make it painful to figure out which bit of the code is misbehaving. Plus, unit tests make it easier to exercise the non-happy paths since you can mock out the homeserver. Codecov is having one of its "special" days, but when it recovers I'd expect it to complain about lack of code coverage. There's a concept called the Test Pyramid which says you should have lots of low-level tests and a few high-level tests.

(Generally, I'd expect to see code, and unit tests that exercise that code, landing in the same commit.)
We should be able to break this PR into smaller chunks, especially if we have more tests. Smaller PRs are generally a good thing since it means that author and reviewer have less state to keep in their head as the PR goes through multiple rounds of review. (Some guy called richvdh wrote a blog post mentioning this.)

crates/matrix-sdk-crypto/src/olm/group_sessions/outbound.rs

richvdh · 2025-08-07T14:34:16Z

crates/matrix-sdk-crypto/src/session_manager/group_sessions/mod.rs

+    ///
+    /// # Panics
+    ///
+    /// Panics if no such session exists for the given room ID, or the session


Suggested change

/// Panics if no such session exists for the given room ID, or the session

/// Panics if no session exists for the given room ID, or the session

richvdh · 2025-08-07T15:00:14Z

crates/matrix-sdk-crypto/src/machine/mod.rs

+    /// * `room_id` - The id of the room for which the message should be
+    ///   encrypted.
+    ///
+    /// * `content` - The plaintext content of the event that should be
+    ///   encrypted as a raw JSON value.
+    ///
+    /// * `state_key` - The associated state key of the event.


event_type is missing here, and the order is inconsistent with the actual arguments

richvdh · 2025-08-07T15:02:41Z

crates/matrix-sdk/src/room/futures.rs

@@ -319,3 +321,83 @@ impl<'a> IntoFuture for SendAttachment<'a> {
        Box::pin(fut.instrument(tracing_span))
    }
 }
+
+/// TODO: Future returned by `Room::send_state_event_raw`.


what does this TODO mean?

richvdh · 2025-08-07T15:05:54Z

crates/matrix-sdk/src/room/futures.rs

+
+/// TODO: Future returned by `Room::send_state_event_raw`.
+#[allow(missing_debug_implementations)]
+pub struct SendStateEventRaw<'a> {


🤔 I wonder if this should be called sendRawStateLikeEvent for consistency with SendRawMessageLikeEvent.

I was going for consistency with Ruma - it uses StateEvent and MessageLikeEvent.

(and Ruma uses "message-like" to disambiguate from m.room.message / m.message)

richvdh · 2025-08-07T15:52:46Z

crates/matrix-sdk/src/test_utils/mod.rs

+        let content =
+            deserialized_event.original_content().expect("The event should not have been redacted");
+
+        assert_matches2::assert_let!($pat = content);


Looks like you've forgotten to use $msg here.

richvdh · 2025-08-07T15:53:26Z

crates/matrix-sdk/src/test_utils/mod.rs

+/// Given a [`TimelineEvent`], assert that the event is a decrypted state
+/// event, and that its content matches the given pattern via a let binding.


Some examples might help me understand how to use this.

richvdh · 2025-08-07T15:55:08Z

crates/matrix-sdk/src/test_utils/mod.rs

+/// event, and that its content matches the given pattern via a let binding.
+#[macro_export]
+macro_rules! assert_let_decrypted_state_event_content {
+    ($pat:pat = $event:expr, $($msg:tt)*) => {


The documentation could do with mentioning that the macro uses the second and subsequent arguments, if provided, as an error message if the content does not match.

richvdh · 2025-08-07T15:59:29Z

testing/matrix-sdk-integration-testing/src/tests/e2ee/state_events.rs

+    // HACK: wait for sync
+    let _ = tokio::time::sleep(Duration::from_secs(3)).await;


We need to find something to wait for, rather than just sleeping for 3 seconds. There's a wait_until_some helper which I imagine you could combine with is_state_encrypted to do what you need.

that said... didn't we agree this was a bug in enable_encryption[_with_state] which should be fixed in a separate PR?

richvdh · 2025-08-07T16:10:17Z

crates/matrix-sdk-crypto/src/store/types.rs

+    /// Whether state event encryption is enabled.
+    pub encrypt_state_events: bool,


RoomSettings is serialized and persisted in clients' local stores. We'll need to make sure it behaves sensibly when the serialized version omits the field. (In other words, needs #[serde(default)] or something)

The commit comment is a bit weird too: it says "WASM SDK" but ... this isn't the wasm bindings project.

TBH I'd suggest pulling this RoomSettings stuff out to a separate PR.

I included WASM SDK in the commit message since this change (afaik) will only affect it and not the rest of the Rust SDK, since it uses the m.room.encryption event to decide whether to encrypt rather than this - I can change?

I think this probably just falls into the more general bracket of "your commit comments could do with a little more detail". You're allowed to use more than one line ;)

I have pulled this out to #5511 - I'll drop the commit that introduces this after our meeting.

…crypt_state to ::encrypt_inner

…elper method

…checks

codspeed-hq · 2025-08-08T15:36:21Z

Unable to generate the performance report

There was an internal error while processing the run's data. We're working on fixing the issue. Feel free to contact us on Discord or at [email protected] if the issue persists.

kaylendog · 2025-08-12T14:22:43Z

Closing in favour of split pull requests:

kaylendog requested review from a team as code owners July 28, 2025 15:06

kaylendog requested review from bnjbvr and richvdh and removed request for a team July 28, 2025 15:06

kaylendog marked this pull request as draft July 28, 2025 15:06

kaylendog self-assigned this Jul 28, 2025

bnjbvr requested review from a team and poljar and removed request for bnjbvr July 28, 2025 15:07

jplatte reviewed Jul 28, 2025

View reviewed changes

crates/matrix-sdk-base/src/room/encryption.rs Outdated Show resolved Hide resolved

kaylendog force-pushed the kaylendog/encrypted-state-events-compat branch 2 times, most recently from 79697a3 to 93feb08 Compare July 30, 2025 13:25

kaylendog force-pushed the kaylendog/encrypted-state-events-compat branch from 9dddf23 to 3dec91a Compare July 31, 2025 15:16

kaylendog force-pushed the kaylendog/encrypted-state-events-compat branch from 3dec91a to ee6a293 Compare August 4, 2025 11:42

kaylendog commented Aug 4, 2025

View reviewed changes

Cargo.toml Outdated Show resolved Hide resolved

richvdh removed their request for review August 4, 2025 16:28

kaylendog force-pushed the kaylendog/encrypted-state-events-compat branch from 4882f4e to 5cb4dd9 Compare August 5, 2025 12:10

kaylendog marked this pull request as ready for review August 5, 2025 12:16

richvdh self-requested a review August 5, 2025 12:57

richvdh reviewed Aug 6, 2025

View reviewed changes

labs/multiverse/src/widgets/room_view/mod.rs Outdated Show resolved Hide resolved

richvdh reviewed Aug 6, 2025

View reviewed changes

testing/matrix-sdk-integration-testing/src/tests/e2ee/state_events.rs Outdated Show resolved Hide resolved

kaylendog force-pushed the kaylendog/encrypted-state-events-compat branch 3 times, most recently from 2c93f9f to 6c49a11 Compare August 6, 2025 16:41

kaylendog force-pushed the kaylendog/encrypted-state-events-compat branch 3 times, most recently from 00c3e99 to b060f96 Compare August 7, 2025 10:57

kaylendog added 16 commits August 7, 2025 15:34

feat(crypto): Add OutboundGroupSession::encrypt_state

aa63ada

Signed-off-by: Skye Elliot <[email protected]>

feat(crypto): Add GroupSessionManager::encrypt_state

2173e36

Signed-off-by: Skye Elliot <[email protected]>

feat(crypto): Add state event encryption methods to OlmMachine

c5fabf5

Signed-off-by: Skye Elliot <[email protected]>

feat(base): Add EncryptionState::StateEncrypted

3192d17

Signed-off-by: Skye Elliot <[email protected]>

feat(sdk): Add SendStateEventRaw

bbf047d

Signed-off-by: Skye Elliot <[email protected]>

feat(sdk): Implement IntoFuture for SendStateEventRaw

7e2c076

Signed-off-by: Skye Elliot <[email protected]>

feat(sdk): Modify Room::send_state_event_raw to return SendStateEventRaw

bde0ef1

Signed-off-by: Skye Elliot <[email protected]>

feat(sdk): Add SendStateEvent

9ef0419

Signed-off-by: Skye Elliot <[email protected]>

feat(sdk): Implement IntoFuture for SendStateEvent

1eef067

Signed-off-by: Skye Elliot <[email protected]>

feat(sdk): Use SendStateEvent future in Room send_state methods

d63f0ee

Signed-off-by: Skye Elliot <[email protected]>

feat(sdk): Add Room::enable_encryption_with_state

89b367f

Signed-off-by: Skye Elliot <[email protected]>

feat(base): Add sync reponse processor for AnySyncState::RoomEncrypted

cd56ffc

Signed-off-by: Skye Elliot <[email protected]>

feat(sdk): Add assert_let_decrypted_state_event_content test utility …

7023b68

…macro Signed-off-by: Skye Elliot <[email protected]>

feat(sdk): Attempt state event decryption in Room::try_decrypt_event

53d3a2e

Signed-off-by: Skye Elliot <[email protected]>

test(e2ee): Add integration test for encrypted state events

a51bec6

Signed-off-by: Skye Elliot <[email protected]>

feat(crypto): Add RoomSettings::enable_encrypted_state_events (WASM SDK)

61d40bd

Signed-off-by: Skye Elliot <[email protected]>

kaylendog force-pushed the kaylendog/encrypted-state-events-compat branch from b060f96 to 61d40bd Compare August 7, 2025 14:34

richvdh requested changes Aug 7, 2025

View reviewed changes

kaylendog added 3 commits August 8, 2025 12:10

refactor(crypto): De-duplicate OutboundGroupSession::encrypt and ::en…

d6e15b6

…crypt_state to ::encrypt_inner

refactor(sdk): De-dupe room::enable_encryption and ...with_state to h…

b707517

…elper method

refactor(sdk): De-dupe SendRawMessageLikeEvent and SendRawStateEvent …

a8b7d11

…checks

kaylendog mentioned this pull request Aug 11, 2025

feat(crypto): Add RoomSettings::encrypt_state_events #5511

Merged

1 task

kaylendog closed this Aug 12, 2025

kaylendog mentioned this pull request Aug 13, 2025

feat: Change type of DecryptedRoomEvent::event to Raw<AnyTimelineEvent> #5512

Merged

2 tasks

kaylendog mentioned this pull request Aug 20, 2025

feat(sdk): Poll encryption state on sync event for up to 3 seconds. #5559

Merged

1 task

	/// Panics if no such session exists for the given room ID, or the session
	/// Panics if no session exists for the given room ID, or the session

		/// Given a [`TimelineEvent`], assert that the event is a decrypted state
		/// event, and that its content matches the given pattern via a let binding.

		// HACK: wait for sync
		let _ = tokio::time::sleep(Duration::from_secs(3)).await;

		/// Whether state event encryption is enabled.
		pub encrypt_state_events: bool,

feat: Encrypted state events (MSC3414) #5456

feat: Encrypted state events (MSC3414) #5456

Uh oh!

Conversation

kaylendog commented Jul 28, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

richvdh commented Jul 28, 2025

Uh oh!

Uh oh!

codspeed-hq bot commented Jul 30, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Merging #5456 will not alter performance

Summary

Uh oh!

zecakeh commented Aug 1, 2025

Uh oh!

Uh oh!

codecov bot commented Aug 4, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

Uh oh!

Uh oh!

richvdh left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

codspeed-hq bot commented Aug 8, 2025

Unable to generate the performance report

Uh oh!

kaylendog commented Aug 12, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Uh oh!

kaylendog commented Jul 28, 2025 •

edited

Loading

codspeed-hq bot commented Jul 30, 2025 •

edited

Loading

codecov bot commented Aug 4, 2025 •

edited

Loading

kaylendog commented Aug 12, 2025 •

edited

Loading