Re: CVE-2013-4287 Algorithmic complexity vulnerability in RubyGems 2.0.7 and older

[Kurt Seifried <kseifried () redhat com>]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 09/17/2013 06:11 PM, Eric Hodel wrote:
> On Sep 16, 2013, at 18:28, Kurt Seifried <kseifried () redhat com>
> wrote:
> > On 09/14/2013 03:11 PM, Alexander Cherepanov wrote:
> > > On 2013-09-10 09:32, Eric Hodel wrote:
> > > > The vulnerability can be fixed by changing the first grouping
> > > > to an atomic grouping in Gem::Version::VERSION_PATTERN in 
> > > > lib/rubygems/version.rb.  For RubyGems 2.0.x:
> > > >
> > > > -  VERSION_PATTERN = 
> > > > '[0-9]+(\.[0-9a-zA-Z]+)*(-[0-9A-Za-z-]+(\.[0-9A-Za-z-]+)*)?'
> > > > # :nodoc: +  VERSION_PATTERN = 
> > > > '[0-9]+(?>\.[0-9a-zA-Z]+)*(-[0-9A-Za-z-]+(\.[0-9A-Za-z-]+)*)?'
> > > > # :nodoc:
> > > >
> > > > For RubyGems 1.8.x:
> > > >
> > > > -  VERSION_PATTERN = '[0-9]+(\.[0-9a-zA-Z]+)*' # :nodoc: + 
> > > > VERSION_PATTERN = '[0-9]+(?>\.[0-9a-zA-Z]+)*' # :nodoc:
> > >
> > > This is not enough. The following script:
> > >
> > > # Regexes are from 
> > > https://github.com/rubygems/rubygems/blob/master/lib/rubygems/version.rb#L150
> >
> > >
VERSION_PATTERN =
> > > '[0-9]+(?>\.[0-9a-zA-Z]+)*(-[0-9A-Za-z-]+(\.[0-9A-Za-z-]+)*)?'
> > > # :nodoc: ANCHORED_VERSION_PATTERN = 
> > > /\A\s*(#{VERSION_PATTERN})*\s*\z/ # :nodoc: 
> > > '1111111111111111111111111111.' =~ ANCHORED_VERSION_PATTERN
> > >
> > > takes ~1m on my machine. The problem is not in VERSION_PATTERN
> > > but in its possible repetition inside
> > > ANCHORED_VERSION_PATTERN.
> >
> > Great, I guess we're going to need a new CVE. Before I assign one
> > can we make sure we fix this so more fiddly expressions don't
> > cause problems? Thanks.
>
> Here's a new patch to go with the new (unassigned) CVE.  This new
> patch replaces regular expression matches that are susceptible to
> backtracking with a parser-like approach.
>
>
>
>
>
> This patch applies to RubyGems 2.1.x releases.  I will create
> patches for RubyGems 1.8.23.1, 1.8.26, 2.0.9 and 2.1.4 if it there
> is no obvious flaw seen in it.
>
> I would like to release this fix by Monday, 23 September as I will
> be traveling mid-week.
>
> The vulnerable regular expression constants are still present, but
> I can't think of a way to construct them that does not allow
> backtracking.  I think they should be removed for the security fix
> release, but a fellow maintainer is worried about backwards
> compatibility and thinks they should be removed in the next feature
> release (2.2).  What do people typically do?
>
> Here is a script to check the patch:

Ok please please use CVE-2013-4363 for this issue (incomplete fix for
CVE-2013-4287).

- -- 
Kurt Seifried Red Hat Security Response Team (SRT)
PGP: 0x5E267993 A90B F995 7350 148F 66BF 7554 160D 4553 5E26 7993
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.14 (GNU/Linux)
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=7el3
-----END PGP SIGNATURE-----