https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=227605

--- Comment #1 from Conrad Meyer <[email protected]> ---
Nice find.  It seems like this can only happen if passed an mbuf chain or uio
longer than crd_len.  Does using the API like this actually make sense?

We could maybe add an assertion that buffer input sizes matches crd_len (+
crd_skip).

In the case of esp_input(), it seems like m_split() could be used to separate
the authentication portion from the encrypted contents before passing to
crypto_dispatch().

OTOH, your patch looks totally correct.  I would additionally cast i to size_t
to avoid ambiguity in comparison signedness in min() macro.

-- 
You are receiving this mail because:
You are the assignee for the bug.
_______________________________________________
[email protected] mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-bugs
To unsubscribe, send any mail to "[email protected]"

Reply via email to