GHSA-rc3f-4j82-cmrg json
Use-after-free of caller-owned server name in client connection constructors
published 17d ago modified 16d ago also RUSTSEC-2026-0413
- cvss
- —
- not scored
- cisa kev
- Not listed
- not in CISA's catalog
- epss
- —
- no score, EPSS covers CVEs only
- public exploits
- 0
- none known
Summary
In rustls-ffi 0.12.0 through 0.15.6, `rustls_client_connection_new()` and `rustls_client_connection_new_alpn()` retained a reference to the caller's `server_name` buffer after returning, rather than copying it. The API documentation did not require the buffer to outlive the call and as a result if the caller freed or reused the buffer, later TLS processing read freed memory.
Clients that free or reuse the `server_name` buffer after the constructor returns are affected. This is most likely for language bindings that pass temporary strings. Callers that keep the `server_name` buffer alive for the lifetime of the `rustls_client_config` are not affected.
Versions 0.12.0 through 0.15.6 are affected. Version 0.15.7+ changed these APIs to copy the input buffer.
Patch / fix
Is GHSA-rc3f-4j82-cmrg in your dependencies? Fencer shows which of your repositories are affected as soon as an advisory lands, and opens the fix as a pull request.
Check with FencerAffected packages
| ecosystem | package | affected | fixed in |
|---|---|---|---|
| crates.io | rustls-ffi | >=0.12.0, <0.15.7 |
0.15.7 |
References
Sources
Fencer Intel keeps one entry per vulnerability: the CVE id wins, then GHSA, and records that alias each other are merged, except that two CVEs always keep separate entries. The same data is available as JSON.