Impact

When ActionDispatch::Cookies reads a signed or encrypted cookie that was written under a rotated secret (config.action_dispatch.cookies_rotations) or in the format that a :hybrid cookie serializer migrates from, it writes the cookie back to the response in the current format, and that write ignores request.session_options[:skip]. ActiveStorage::DisableSession relies on that option to keep the session cookie off the public, immutable responses of the Active Storage proxy controllers, but those controllers still load the session, so a request carrying an old-format session cookie could receive the rewritten session cookie in Set-Cookie on a cacheable response. A shared cache that stores such a response together with its Set-Cookie header could then serve one user's session to later requesters of the same URL, or serve an attacker's session to other users.

Applications are affected when they serve Active Storage files through the proxy controllers, have a cookie rotation or a :hybrid serializer migration configured, and run behind a cache that stores Set-Cookie on public responses. This is a path around the fix for CVE-2024-26144.

Details

In the fixed releases, a response that skips the session leaves an old-format signed or encrypted cookie unchanged, and the next request that does not skip the session upgrades it. An application that skips the session on every request no longer upgrades its signed or encrypted cookies. Cookies that the application sets explicitly are still written. See the Action Pack CHANGELOG.

Workarounds

Configure caching proxies not to store responses that carry a Set-Cookie header, or complete the cookie rotation or serializer migration and remove it from the configuration.

Releases

The fixed releases are available at the normal locations.

Patches for those who cannot upgrade at once:

• 8.1: https://github.com/rails/rails/commit/849a6c30eacbb6ca4722132f43c4333a93188556.patch
• 8.0: https://github.com/rails/rails/commit/0e6cf8faa89caf4c64311d35ca2bc88c1458d4e3.patch

Credit

This issue was responsibly reported by HackerOne researcher flashmyname.