Try to load keys from key backup when a message fails to decrypt #2373
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
turt2live
reviewed
May 16, 2022
t3chguy
reviewed
May 17, 2022
Co-authored-by: Travis Ralston <travisr@matrix.org>
Co-authored-by: Travis Ralston <travisr@matrix.org>
Co-authored-by: Travis Ralston <travisr@matrix.org>
Co-authored-by: Travis Ralston <travisr@matrix.org>
| @@ -1298,7 +1298,9 @@ class MegolmDecryption extends DecryptionAlgorithm { | |||
|
|
|||
| if (res === null) { | |||
| // We've got a message for a session we don't have. | |||
| // | |||
| // try and get the missing key from the backup first | |||
| this.crypto.backupManager.queryKeyBackupRateLimited(event.getRoomId(), content.session_id).catch(() => {}); | |||
There was a problem hiding this comment.
In order to reduce unneeded to_device traffic, on EA if we have a positive reply from the backup we don't call requestKeysForEvent. Is it possible to do the same here?
There was a problem hiding this comment.
Not trivially. If we naively await this operation, we get stuck in a deadlock because the key backup query operation includes importing the keys, which includes re-attempting decryption of the messages, which means waiting for the current decryption attempt to finish. So we can't use, or even know, the results of the backup query during the current decryption attempt, without a bunch of messy re-architecting. In order to still avoid sending key requests, we'd need to add more complicated state tracking to check whether we've already tried the key backups on a previous decryption attempt, as well as logic to immediately queue another decryption attempt if the backup query fails.
turt2live
approved these changes
May 31, 2022
su-ex
added a commit
to SchildiChat/matrix-js-sdk
that referenced
this pull request
Jul 7, 2022
* Remove unused sessionStore ([\matrix-org#2455](matrix-org#2455)). * Implement MSC3827: Filtering of `/publicRooms` by room type ([\matrix-org#2469](matrix-org#2469)). * expose latestLocationEvent on beacon model ([\matrix-org#2467](matrix-org#2467)). Contributed by @kerryarchibald. * Live location share - add start time leniency ([\matrix-org#2465](matrix-org#2465)). Contributed by @kerryarchibald. * Log real errors and not just their messages, traces are useful ([\matrix-org#2464](matrix-org#2464)). * Various changes to `src/crypto` files for correctness ([\matrix-org#2137](matrix-org#2137)). Contributed by @ShadowJonathan. * Update MSC3786 implementation: Check the `state_key` ([\matrix-org#2429](matrix-org#2429)). * Timeline needs to refresh when we see a MSC2716 marker event ([\matrix-org#2299](matrix-org#2299)). Contributed by @MadLittleMods. * Try to load keys from key backup when a message fails to decrypt ([\matrix-org#2373](matrix-org#2373)). Fixes element-hq/element-web#21026. Contributed by @duxovni. * Send call version `1` as a string ([\matrix-org#2471](matrix-org#2471)). Fixes element-hq/element-web#22629. * Fix issue with `getEventTimeline` returning undefined for thread roots in main timeline ([\matrix-org#2454](matrix-org#2454)). Fixes element-hq/element-web#22539. * Add missing `type` property on `IAuthData` ([\matrix-org#2463](matrix-org#2463)). * Clearly indicate that `lastReply` on a Thread can return falsy ([\matrix-org#2462](matrix-org#2462)). * Fix issues with getEventTimeline and thread roots ([\matrix-org#2444](matrix-org#2444)). Fixes element-hq/element-web#21613. * Live location sharing - monitor liveness of beacons yet to start ([\matrix-org#2437](matrix-org#2437)). Contributed by @kerryarchibald. * Refactor Relations to not be per-EventTimelineSet ([\matrix-org#2412](matrix-org#2412)). Fixes matrix-org#2399 and element-hq/element-web#22298. * Add tests for sendEvent threadId handling ([\matrix-org#2435](matrix-org#2435)). Fixes element-hq/element-web#22433. * Make sure `encryptAndSendKeysToDevices` assumes devices are unique per-user. ([\matrix-org#2136](matrix-org#2136)). Fixes matrix-org#2135. Contributed by @ShadowJonathan. * Don't bug the user while re-checking key backups after decryption failures ([\matrix-org#2430](matrix-org#2430)). Fixes element-hq/element-web#22416. Contributed by @duxovni.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Fixes element-hq/element-web#21026.
Also typescriptified
TestClientwhile I was in the area, I can split that into a separate PR if that'd be preferable.Here's what your changelog entry will look like:
✨ Features