full-disclosure - Re: Internet Explorer Crash

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <6128.1176915351@turing-police.cc.vt.edu>
Date: Wed, 18 Apr 2007 12:55:51 -0400
From: Valdis.Kletnieks@...edu
To: admin@...ibase.ca
Cc: full-disclosure@...ts.grok.org.uk
Subject: Re: Internet Explorer Crash

On Wed, 18 Apr 2007 12:31:57 EDT, Kradorex Xeron said:
> There should be an implimentation on ALL browsers that a loop such large is 
> unacceptable and refuse to even run it. There is no viable reason for a 
> client-side to run a loop through so many itterations.

There's this thing called the Turing Halting Problem. :)

The problem is that it's *really* hard to *programatically* look at a
loop like that, and say "That's going to loop 'too long' (for some fuzzy
definition of 'too long')".  Take that same code, and change the comparison
to 'while (z.length < 3)'.  Does that loop "too long"?  How about '< 8'?
(Keep in mind that to check this *from within*, it needs to have the knowledge
that z is the loop control, which it has, and that z.length is approximately
log10(z), and that some value of log10(z) is "too much".  And once you've coded
all that knowledge, the attacker just changes the test to this:

   while (foo(z) == TRUE) (...

and foo(z) is defined as:

boolean foo(int z) {
     static a = 0;
     if (isprime(z)) a++;
     if a > 1000000 return FALSE;
     return TRUE;
}

Bonus points for defining isprime() as "Sieve of Eratosthenes" rather than some
higher-performance primality check like Rabin-Miller or similar.  Or maybe not
- Sieve is probably simple enough that you can special-case it, better methods
have more obscure internals. And we're *trying* to burn CPU - so maybe "Sieve
of Eratosthene's less clever brother" is called for (iterate 1 to N, rather
than 1 to sqrt(N)) :)

So - other than "it has already burned more than N seconds of CPU", what
test do you propose to make?  And what do you do if the site is some
Javascript-driven interface to a corporate application that the user is
expected to be in all day, and it's *legitimate* to burn lots more than N
seconds during an 8-hour day?

(Hint - "trusted site" is probably not the greatest way to phrase that sort
of check... ;)

Content of type "application/pgp-signature" skipped

_______________________________________________
Full-Disclosure - We believe in it.
Charter: http://lists.grok.org.uk/full-disclosure-charter.html
Hosted and sponsored by Secunia - http://secunia.com/