Re: [TLS] [oss-security] CVE-2009-3555 for TLS renegotiation MITM attacks

[Marsh Ray <marsh at extendedsubset.com>]

ArkanoiD wrote:
> BTW renegotiation handshake looks quite similar to initial handshake from
> the client point of view;

Yeah, currently it looks identical.

> is there a way to detect the attack on client side
> and drop the connection?

With draft-rescorla-tls-renegotiate, the new header comes back to the
client on Server Hello, and this notifies the client that the server
believes it's a re-negotation. Even better, it ties it strongly to the
previous session, so legit renegotiations are protected, too.

- Marsh