5 ms·

The POODLE bites again

• huxley 12y ago
  Link didn't work for me, here is the Google Cache text version: http://webcache.googleusercontent.com/search?q=cache:f01MHrX8LTAJ:https://www.imperialviolet.org/2014/12/08/poodleagain.html&strip=1 http://webcache.googleusercontent.com/search?q=cache:f01MHrX...
• resonantcore 12y ago
  We look forward to TLS 1.2 support being the norm. (And then, hopefully, the ratification and adoption of TLS 1.3) A 50% adoption rate is excellent news. Still a long way to go, but that's worth toasting over.
  • yuhong 12y ago
    What is frustrating is how many such servers have TLS 1.3 intolerance (even PayPal), and often the same servers are also affected by this bug. I wonder what TLS implementation is this.
    • feld 12y ago
      TLS 1.3 isn't finished yet...
      • yuhong 12y ago
        I know, the goal is to prepare.
        • feld 12y ago
          I'm not sure what you mean then by servers having TLS 1.3 intolerance.
          • yuhong 12y ago
            It does not respond properly to clients trying to negotiate TLS 1.3. It should return a ServerHello showing the supported TLS 1.2 version.
            • feld 12y ago
              I wasn't aware there was a TLS 1.3 PoC that people could use for testing.
              • yuhong 12y ago
                It is as simple as setting the version in the ClientHello message.
                • feld 12y ago
                  Oh, I see
• eyeareque 12y ago
  Are there any more details than what this write up contains?
• cryptbe 12y ago
  POODLE worked not only against SSLv3, but also against any TLS implementations that check padding in SSLv3's style (e.g., just checking the last byte, and ignoring the rest of the padding). SSL accelerators from F5 and A10 were vulnerable. Thus, many of the world's largest sites were vulnerable.
• tacoman 12y ago
  "This seems like a good moment to reiterate that everything less than TLS 1.2 with an AEAD cipher suite is cryptographically broken." So this means AES-GCM essentially?
  • colmmacc 12y ago
    Well, almost. AES-GCM is also "Cryptographically Broken" if you take into account software implementations using table lookup schemes that are subject to cache timing attacks. But the fundamental AEAD support in TLS1.2 is a much better place to be overall - and GCM tags are much better than the MAC-then-Encrypt HMACs elsewhere in TLS. For practical purposes; AES-GCM is the "least worst", by a long way.
    • cesarb 12y ago
      For the future, there's also the encrypt-then-MAC extension (https://tools.ietf.org/html/rfc7366 https://tools.ietf.org/html/rfc7366, https://bugzilla.mozilla.org/show_bug.cgi?id=972145 https://bugzilla.mozilla.org/show_bug.cgi?id=972145), and probably some variant of a ChaCha20+Poly1305 AEAD (https://tools.ietf.org/html/draft-agl-tls-chacha20poly1305-04 https://tools.ietf.org/html/draft-agl-tls-chacha20poly1305-0..., https://bugzilla.mozilla.org/show_bug.cgi?id=917571 https://bugzilla.mozilla.org/show_bug.cgi?id=917571, already on Chrome AFAIK). For now, AES-GCM is the best alternative.
      • edwintorok 12y ago
        I like the alternative described here: https://www.imperialviolet.org/2014/02/27/tlssymmetriccrypto.html https://www.imperialviolet.org/2014/02/27/tlssymmetriccrypto... "The end result is that Chrome talking to Google uses AES-GCM if there's hardware support at the client and ChaCha20-Poly1305 otherwise." It seems to be specific to Chrome though, and all TLS clients would have to reimplement that choice. Would be good if there was a way to tell the SSL library to give you the best cipher that works on your hardware (i.e. don't give AES-GCM/AES-CBC when there is no hardware support and the software implementation isn't constant time).