http://qs321.pair.com?node_id=1034695


in reply to Re^7: 5.18.0 is available NOW!
in thread 5.18.0 is available NOW!

Two problems with that retort:

  1. It would be hard to code a patch to handle an attack vector that -- to the best of my ability to discover; and despite requests for further information and a promise of "I would release a full-disclosure document in the middle to last week of march." -- it seems has never been publicly described, let alone demonstrated.

    Indeed -- whilst I'm still waiting to hear back from mitre (CVE DB maintainers) and a couple of other likely organisations -- I can find no trace that anyone other than demerphq has ever been made party to the details of the vulnerability.

  2. Also, based upon the scant information I have been able to glean -- and a lot of unfortunately necessary supposition -- it seems likely that any one of several one-line patches might serve to totally mitigate the possibility of CVE-2013-1667.

    With the added upside that almost none of the pain caused by the implemented solution would have been necessary.

I'm preparing a paper -- which will probably come in 4 or 5 parts -- now. But it would surely be easier, and maybe even unnecessary, if disclosure were made.

With the rise and rise of 'Social' network sites: 'Computers are making people easier to use everyday'
Examine what is said, not who speaks -- Silence betokens consent -- Love the truth but pardon error.
"Science is about questioning the status quo. Questioning authority".
In the absence of evidence, opinion is indistinguishable from prejudice.

Replies are listed 'Best First'.
Re^9: 5.18.0 is available NOW!
by demerphq (Chancellor) on May 22, 2013 at 10:13 UTC

    It turns out that vendors are slow in providing updates. We did a survey and there were too many vulnerable systems to release the details of the attack. Once we feel most of the affected systems are patched we will release more details. This process is called "responsible disclosure".

    Also, based upon the scant information I have been able to glean -- and a lot of unfortunately necessary supposition -- it seems likely that any one of several one-line patches might serve to totally mitigate the possibility of CVE-2013-1667.

    The patches mitigating CVE-2013-1667 are all public. The patches which changed Perls hash implementation are all public. The only code which is not public is the code which demonstrates a key-discovery attack on perls old hash function, and the key generator code to produce an attack key set for CVE-2013-1667.

    Please stop posting FUD about this issue. You do not know what you are talking about, and everybody reading this thread should know it.

    ---
    $world=~s/war/peace/g

[reply]
[d/l]
      This process is called "responsible disclosure".

      That can lead to "a feeling of false security."

      The patches mitigating CVE-2013-1667 are all public.

      The patches are public; but whether they actually address the perceived problem -- nor even if the perceived problem is actually a problem -- cannot be determined without knowing what the problem is.

      The only code which is not public is the code which demonstrates a key-discovery attack on perls old hash function,

      Easily reproduced with a 20 line script. It is running now:

      <Reveal this spoiler or all in this thread>
      You do not know what you are talking about

      Actually, I do. As you will find out.

      With the rise and rise of 'Social' network sites: 'Computers are making people easier to use everyday'
      Examine what is said, not who speaks -- Silence betokens consent -- Love the truth but pardon error.
      "Science is about questioning the status quo. Questioning authority".
      In the absence of evidence, opinion is indistinguishable from prejudice.
[reply]
[d/l]

        The patches are public; but whether they actually address the perceived problem -- nor even if the perceived problem is actually a problem -- cannot be determined without knowing what the problem is.

        Actually, I do. As you will find out.

        I look forward to reading your patch. But I find it hard to reconcile your two statements. On one hand you admit you don't know what the problem is, yet on the other you assert you know what you are talking about. Both can't be right. IMO if you understand the patches you should be able to figure out the attack and recreate it.

        ---
        $world=~s/war/peace/g

[reply]

In Section Perl News