client: reject TLS 1.3 compat session ID in 1.2 #2360
Merged
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
cpu
commented
Feb 26, 2025
| @@ -17,7 +17,6 @@ | |||
| "CheckLeafCurve": "", | |||
| "SendWarningAlerts-*": "", | |||
| "Peek-*": "", | |||
| "EchoTLS13CompatibilitySessionID": "", | |||
There was a problem hiding this comment.
It looks like this ignore was added way back in 2018 (32eeec6) without any special mention.
I was looking at the test case for other reasons and couldn't think of a reason why we shouldn't add the coverage. WDYT?
cpu
commented
Feb 26, 2025
rustls/src/error.rs
Outdated
| @@ -281,6 +281,7 @@ pub enum PeerMisbehaved { | |||
| UnsolicitedServerHelloExtension, | |||
| WrongGroupForKeyShare, | |||
| UnsolicitedEchExtension, | |||
| ServerEchoedCompatibilitySessionId, | |||
There was a problem hiding this comment.
I initially added this in alpha-order, but the semver compat check job doesn't like a new value being sliced in even though this is non-exhaustive because:
This breaks downstream code that used its value via a numeric cast like
as isize.
There was a problem hiding this comment.
Pff, not sure we should listen to it on that.
There was a problem hiding this comment.
I put it back in sorted position for now and can revisit if someone feels strongly about maintaining the numeric values.
|
Cool -- how did you find this? |
djc
approved these changes
Feb 26, 2025
Codecov ReportAll modified and coverable lines are covered by tests ✅
Additional details and impacted files@@ Coverage Diff @@
## main #2360 +/- ##
=======================================
Coverage 94.89% 94.89%
=======================================
Files 103 103
Lines 24532 24548 +16
=======================================
+ Hits 23280 23296 +16
Misses 1252 1252 ☔ View full report in Codecov by Sentry. |
In $JOB_HOURS I'm working through the Go standard library's equivalent BoGo skip list trying to improve coverage. I have a similar change open there after thinking through the testcase. Where possible I've been trying to (in volunteer time) bring useful changes like this and #2347 back here 🤝 |
Benchmark resultsInstruction countsSignificant differencesThere are no significant instruction count differences Other differencesClick to expand
Wall-timeSignificant differencesThere are no significant wall-time differences Other differencesClick to expand
Additional informationCheckout details:
|
If the client hello input had no resuming session, but has a non-empty SessionId, it means we generated a TLS 1.3 legacy_session_id for middle-box compatibility purposes. In this case if we end up handshaking for TLS 1.2 we should ensure the server didn't echo the session ID back to us. Since it was invented at random to fill the need of a non-empty value it should never be a recognized server session.
51eef44 to
c414bfc
Compare
ctz
approved these changes
Feb 27, 2025
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.
If the client hello input had no resuming session, but has a non-empty SessionId, it means we generated a TLS 1.3 legacy_session_id for middle-box compatibility purposes.
In this case if we end up handshaking for TLS 1.2 we should ensure the server didn't echo the session ID back to us. Since it was invented at random to fill the need of a non-empty value it should never be a recognized server session.