crypto: rewrite check_sender_trust_requirement to use VerificationState
#5044
Conversation
This file contains hidden or 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
Codecov ReportAttention: Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## main #5044 +/- ##
=======================================
Coverage 85.85% 85.85%
=======================================
Files 325 325
Lines 35936 35938 +2
=======================================
+ Hits 30853 30855 +2
Misses 5083 5083 ☔ View full report in Codecov by Sentry. 🚀 New features to boost your workflow:
|
richvdh
force-pushed
the
rav/refactor_check_sender_trust_requirement
branch
from
727b356 to
2a4b503
Compare
richvdh
changed the title
check_sender_trust_requirement to use VerficationStatecheck_sender_trust_requirement to use VerificationState
…tate` There are two reasons for this. Firstly. we've already done a bunch of work to map `SenderData` into a `VerificationState`, and the decision tree from `VerificationState` to allow/reject is simpler than going from `SenderData`, even if we have to fudge it a bit to get the "legacy" flag. (Note that it allows us to get rid of an `unreachable!` panic.) Secondly, `VerficationState` represents the state of an *event*, whereas `SenderData` is about the session as a whole. A session can be fine, whilst event (claiming to be) encrypted with it can be suspect. What we want here is to check a specific message. Currently, this doesn't make any functional difference, but conceptually it's cleaner to check the `VerificationState`.
richvdh
force-pushed
the
rav/refactor_check_sender_trust_requirement
branch
from
2a4b503 to
ad99431
Compare
richvdh
commented
May 15, 2025
Comment on lines
1827
to
1832
| matches!( | ||
| (verification_level, legacy_session), | ||
| (VerificationLevel::UnverifiedIdentity, _) | ||
| | (VerificationLevel::UnsignedDevice, true) | ||
| | (VerificationLevel::None(_), true) | ||
| ) |
There was a problem hiding this comment.
correspondances between VerificationLevel and SenderData:
SenderData::VerificationViolation -> VerificationLevel::VerificationViolation
SenderData::SenderUnverified -> VerificationLevel::UnverifiedIdentity
SenderData::DeviceInfo -> VerificationLevel::UnsignedDevice
SenderData::UnknownDevice -> VerificationLevel::None
There is no equivalent to SenderData::SenderVerified; that is handled as VerificationState::Verified above.
richvdh
requested review from
stefanceriu and
poljar
and removed request for
a team
poljar
approved these changes
May 16, 2025
| // respectively, and those cases may be acceptable if the reason | ||
| // for the lack of data is that the sessions were established | ||
| // before we started collecting SenderData. | ||
| let legacy_session = match &session.sender_data { |
There was a problem hiding this comment.
Suggested change
| let legacy_session = match &session.sender_data { | |
| let legacy_session = match session.sender_data { |
| let legacy_session = match &session.sender_data { | ||
| SenderData::DeviceInfo { legacy_session, .. } => legacy_session, | ||
| SenderData::UnknownDevice { legacy_session, .. } => legacy_session, | ||
| _ => &false, |
There was a problem hiding this comment.
Suggested change
| _ => &false, | |
| _ => false, |
Comment on lines
1835
to
1836
| TrustRequirement::CrossSigned => { | ||
| matches!(verification_level, VerificationLevel::UnverifiedIdentity) |
There was a problem hiding this comment.
Again, listing all cases makes sense to me:
// If cross-signing of identities is required, the only acceptable unverified case
// is when the identity is signed but not yet verified by us.
TrustRequirement::CrossSigned => match verification_level {
VerificationLevel::UnverifiedIdentity => true,
VerificationLevel::VerificationViolation
| VerificationLevel::UnsignedDevice
| VerificationLevel::None(_) => false,
},
Comment on lines
1827
to
1832
| matches!( | ||
| (verification_level, legacy_session), | ||
| (VerificationLevel::UnverifiedIdentity, _) | ||
| | (VerificationLevel::UnsignedDevice, true) | ||
| | (VerificationLevel::None(_), true) | ||
| ) |
There was a problem hiding this comment.
I don't like using matches!() in crypto related code, it's very easy to forget to modify the statement once new variants are added to the enum. I know it's concise and looks neat, it still makes me worry for the future.
What do you think about listing it all out like so:
// In the CrossSignedOrLegacy case the following rules apply:
//
// 1. Identities we have not yet verified can be decrypted regardless of the
// legacy state of the session.
// 2. Devices that aren't signed by the owning identity of the device can only
// be decrypted if it's a legacy session.
// 3. If we have no information about the device, we should only decrypt if it's
// a legacy session.
// 4. Anything else, should throw an error.
match (verification_level, legacy_session) {
// Case 1
(VerificationLevel::UnverifiedIdentity, _) => true,
// Case 2
(VerificationLevel::UnsignedDevice, true) => true,
// Case 3
(VerificationLevel::None(_), true) => true,
// Case 4
(VerificationLevel::VerificationViolation, _)
| (VerificationLevel::UnsignedDevice, _)
| (VerificationLevel::None(_), false) => false,
}There was a problem hiding this comment.
I agree it's clearer. I started with that, but I got a lint failure. I'll disable the lint, I guess.
…nder_trust_requirement
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.
There are two reasons for this.
Firstly. we've already done a bunch of work to map
SenderDatainto aVerificationState, and the decision tree fromVerificationStateto allow/reject is simpler than going fromSenderData, even if we have to fudge it a bit to get the "legacy" flag. (Note that it allows us to get rid of anunreachable!panic.)Secondly,
VerficationStaterepresents the state of an event, whereasSenderDatais about the session as a whole. A session can be fine, whilst events (claiming to be) encrypted with it can be suspect. What we want here is to check a specific message. Currently, this doesn't make any functional difference, but conceptually it's cleaner to check theVerificationState.Note that there are a bunch of tests for this method in
matrix-sdk-crypto/src/machine/tests/decryption_verification_state.rs, calledtest_decryption_trust_requirement.