* Restructure eventsPendingKey to remove sender key
For withheld notices, we don't necessarily receive the sender key, so we'll
jhave to do without it.
* Re-decrypt events when we receive a withheld notice
* Extend test to cover late-arriving withheld notices
* update unit tests
* Inline subtlecrypto shim
The presence of this thing just makes code more confusing.
* Remove pre-node-20 webcrypto hack
Until node 20.0, the webcrypto API lived at `crypto.webCrypto`. It's now
available at the same place as in web -- `globalThis.crypto`.
See: https://nodejs.org/docs/latest-v20.x/api/webcrypto.html#web-crypto-api
* oidc auth test: Clean up mocking
THe previous reset code wasn't really resetting the right thing. Let's just
re-init `window.crypto` on each test.
* Remove `crypto` shim
This isn't very useful any more.
We had both an `onIncomingKeyVerificationRequest` and an
`onKeyVerificationRequest` which did different, but related, things.
Improve the documentation and reduce the duplication.
* Use server name instead of homeserver url to allow well-known lookups during QR OIDC reciprocation
Signed-off-by: Michael Telatynski <7t3chguy@gmail.com>
* Iterate
Signed-off-by: Michael Telatynski <7t3chguy@gmail.com>
* Iterate
Signed-off-by: Michael Telatynski <7t3chguy@gmail.com>
* Iterate
Signed-off-by: Michael Telatynski <7t3chguy@gmail.com>
* Iterate
Signed-off-by: Michael Telatynski <7t3chguy@gmail.com>
---------
Signed-off-by: Michael Telatynski <7t3chguy@gmail.com>
* RustCrypto: Move CryptoBackend impl to CryptoBackend impl section
Given there is a `CryptoBackend implementation` section, the methods
implementing CryptoBackend should be there.
* RustCrypto: Fix documentation on dehydration methods
* RustCrypto: reunite `resetKeyBackup` with its helper
A couple of new methods had snuck into the middle.
* Add crypto methods for OIDC QR code login
Signed-off-by: Michael Telatynski <7t3chguy@gmail.com>
* Improve test
Signed-off-by: Michael Telatynski <7t3chguy@gmail.com>
* Revert test due to hang inside Rust.
Signed-off-by: Michael Telatynski <7t3chguy@gmail.com>
* Iterate
Signed-off-by: Michael Telatynski <7t3chguy@gmail.com>
* Update test name
Signed-off-by: Michael Telatynski <7t3chguy@gmail.com>
* Update test name
Signed-off-by: Michael Telatynski <7t3chguy@gmail.com>
---------
Signed-off-by: Michael Telatynski <7t3chguy@gmail.com>
Slightly more involved than normal because it requires us to pass a backup version into OlmMachine.importBackedUpRoomKeys.
On the other hand we can now re-enable the test that was disabled in #4214 due to matrix-org/matrix-rust-sdk#3447Fixes: element-hq/element-web#27165
* Add `PerSessionKeyBackupDownloader.isKeyBackupDownloadConfigured()`
* Add new `RustBackupManager.getServerBackupInfo`
... and a convenience method in PerSessionKeyBackupDownloader to access it.
* Crypto.spec: move `useRealTimers` to global `afterEach`
... so that we don't need to remember to do it everywhere.
* Use fake timers for UTD error code tests
This doesn't have any effect on the tests, but *does* stop jest from hanging
when you run the tests in in-band mode. It shouldn't *really* be needed, but
using fake timers gives more reproducible tests, and I don't have the
time/patience to debug why it is needed.
* Use new error codes for UTDs from historical events
* initial implementation of device dehydration
* add dehydrated flag for devices
* add missing dehydration.ts file, add test, add function to schedule dehydration
* add more dehydration utility functions
* stop scheduled dehydration when crypto stops
* bump matrix-crypto-sdk-wasm version, and fix tests
* adding dehydratedDevices member to mock OlmDevice isn't necessary any more
* fix yarn lock file
* more tests
* fix test
* more tests
* fix typo
* fix logic for checking if dehydration supported
* make changes from review
* add missing file
* move setup into another function
* apply changes from review
* implement simpler API
* fix type and move the code to the right spot
* apply suggestions from review
* make sure that cross-signing and secret storage are set up
* Clean up decryption failure integ tests
* Fix the names
* Stop waiting as soon as the event is decrypted, even if code is wrong (so
tests fail rather than time out if the code is wrong)
* Bump timeouts on some tests
These tend to fail due to slow init of wasm artifacts
* Factor out `onDecryptionKeyMissingError` call
* Factor out `onMegolmDecryptionError`
* crypto.spec: make `keyResponder` a local var
it is never used between functions, so making it external was confusing
* Persist encryption state to the rust room list.
* `MatrixClient.shouldEncryptEventForRoom`: fix for rust crypto
Previously, we were not bothering to ask the Rust Crypto stack if it thought we
should be encrypting for a given room. This adds a new method to `CryptoApi`,
wires it up for legacy and Rust crypto, and calls it.
* Tests for persistent room list
* `getOwnDeviceKeys`: use `olmMachine.identityKeys`
This is simpler, and doesn't rely on us having done a device query to work.
* Factor out `requestKeyBackupVersion` utility
* Factor out `makeMatrixHttpApi` function
* Convert `initRustCrypto` to take a params object
* Improve logging in startup
... to help figure out what is taking so long.
* initial commit
* new interation test
* more comments
* fix test, quick refactor on request version
* cleaning and logs
* fix type
* cleaning
* remove delegate stuff
* remove events and use timer mocks
* fix import
* ts ignore in tests
* Quick cleaning
* code review
* Use Errors instead of Results
* cleaning
* review
* remove forceCheck as not useful
* bad naming
* inline pauseLoop
* mark as paused in finally
* code review
* post merge fix
* rename KeyDownloadRateLimit
* use same config in loop and pass along
* don't back up keys that we got from backup
* lint
* lint again
* remove key source struct and add function for importing from backup
* apply changes from review
---------
Co-authored-by: Richard van der Hoff <1389908+richvdh@users.noreply.github.com>
* Rust-crypto: fix `bootstrapCrossSigning` on second call
Currently, `bootstrapCrossSigning` raises an exception if it is called a second
time before secret storage is set up. It is easily fixed by checking that 4S is
set up before trying to export to 4S.
Also a few logging fixes while we're in the area.
* Factor out an `AccountDataAccumulator`
* Another test for bootstrapCrossSigning
* Bump `matrix-rust-sdk-crypto-wasm` to v3.2.0
* Reinstate timeout on `getUserDevices` call
Turns out that this used to have a timeout of 1 second in the wasm
bindings, which it no longer does. Reinstate it here.
MatrixClient API was exposing two methods that only worked for legacy crypto:
- getDeviceEd25519Key
- getDeviceCurve25519Key
=> These are used in the react-sdk for some functionality (rageshake, sentry, rendez-vous).
I have deprecated those calls from MatrixClient and created a new API in CryptoApi (where it belongs):
getOwnDeviceKeys(): Promise<OwnDeviceKeys>
If we call methods on `OlmMachine` after `MatrixClient.stopClient` is called,
we will end up with a "use of moved value" error. We can turn these into
something more useful with judicious use of `getOlmMachineOrThrow`.
Alternatively, we can sidestep the issue by bailing out sooner.
* Reduce console log spam
A couple of different things:
* Increase the `MaxListeners` setting on `MatrixClient` and `Thread`, so that
we don't get "possible EventEmitter leak" warnings
* Disable a couple of warnings/info lines that are just part of regular
operation and are logged in large volumes.
* another noisy log line
* Reinstate warning about receipts for missing events
Apparently this is being worked on
* fix members loaded on intitial sync
* Update test to use KeyResponder
* Use E2EKeyResponder
* code review
* better comment
* fix test
* post merge fix
* fix imports
* refactoring, better names
* code review
* clean tests
* Cleanups per review comments
* fix test
* Apply suggestions from code review
---------
Co-authored-by: Richard van der Hoff <richard@matrix.org>
Co-authored-by: Richard van der Hoff <1389908+richvdh@users.noreply.github.com>
* Element-R: reduce log spam when checking server key backup
Fixes a lot of spam in the logs about "uncaught in promise: No room_keys
found".
* Improve integ tests for backup query after UTD
* Yield in the backup decryption loop
* Fix another broken test
The current deprecation notice advises you to use a method which does something
completely different.
Fixing this "properly" is slightly challenging because we don't want to support
setting it to `true` in Rust Crypto; yet I don't really want to change the
default for legacy crypto.
Let's just document the behaviour for now.
* Add current version of the rust-sdk and vodozemac
* Return OlmVersion in `CryptoApi#getVersion` for old crypto
* Add `Olm` prefix
* Fix documentation
* Review changes
