Contents of /xml/htdocs/proj/en/glep/glep-0059.html

Parent Directory Parent Directory | Revision Log Revision Log

Revision 1.2 - (show annotations) (download) (as text)
Wed Oct 22 18:03:40 2008 UTC (5 years, 5 months ago) by robbat2
Branch: MAIN
Changes since 1.1: +8 -14 lines
File MIME type: text/html
Update HTML to match RST validation of tree-signing fixes.

1 <?xml version="1.0" encoding="utf-8" ?>
2 <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
3 <html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en" lang="en">
5 <head>
6 <meta http-equiv="Content-Type" content="text/html; charset=utf-8" />
7 <meta name="generator" content="Docutils 0.5: http://docutils.sourceforge.net/" />
8 <title>GLEP 59 -- Manifest2 hash policies and security implications</title>
9 <link rel="stylesheet" href="tools/glep.css" type="text/css" /></head>
10 <body bgcolor="white">
11 <table class="navigation" cellpadding="0" cellspacing="0"
12 width="100%" border="0">
13 <tr><td class="navicon" width="150" height="35">
14 <a href="http://www.gentoo.org/" title="Gentoo Linux Home Page">
15 <img src="http://www.gentoo.org/images/gentoo-new.gif" alt="[Gentoo]"
16 border="0" width="150" height="35" /></a></td>
17 <td class="textlinks" align="left">
18 [<b><a href="http://www.gentoo.org/">Gentoo Linux Home</a></b>]
19 [<b><a href="http://www.gentoo.org/proj/en/glep">GLEP Index</a></b>]
20 [<b><a href="http://www.gentoo.org/proj/en/glep/glep-0059.txt">GLEP Source</a></b>]
21 </td></tr></table>
22 <table class="rfc2822 docutils field-list" frame="void" rules="none">
23 <col class="field-name" />
24 <col class="field-body" />
25 <tbody valign="top">
26 <tr class="field"><th class="field-name">GLEP:</th><td class="field-body">59</td>
27 </tr>
28 <tr class="field"><th class="field-name">Title:</th><td class="field-body">Manifest2 hash policies and security implications</td>
29 </tr>
30 <tr class="field"><th class="field-name">Version:</th><td class="field-body">1.2</td>
31 </tr>
32 <tr class="field"><th class="field-name">Last-Modified:</th><td class="field-body"><a class="reference external" href="http://www.gentoo.org/cgi-bin/viewcvs.cgi/xml/htdocs/proj/en/glep/glep-0059.txt?cvsroot=gentoo">2008/10/22 17:59:43</a></td>
33 </tr>
34 <tr class="field"><th class="field-name">Author:</th><td class="field-body">Robin Hugh Johnson &lt;robbat2&#32;&#97;t&#32;gentoo.org&gt;,</td>
35 </tr>
36 <tr class="field"><th class="field-name">Status:</th><td class="field-body">Draft</td>
37 </tr>
38 <tr class="field"><th class="field-name">Type:</th><td class="field-body">Standards Track</td>
39 </tr>
40 <tr class="field"><th class="field-name">Content-Type:</th><td class="field-body"><a class="reference external" href="glep-0002.html">text/x-rst</a></td>
41 </tr>
42 <tr class="field"><th class="field-name">Requires:</th><td class="field-body"><a class="reference external" href="http://www.gentoo.org/proj/en/glepglep-0044.html">44</a></td>
43 </tr>
44 <tr class="field"><th class="field-name">Created:</th><td class="field-body">October 2006</td>
45 </tr>
46 <tr class="field"><th class="field-name">Updated:</th><td class="field-body">November 2007, June 2008, July 2008</td>
47 </tr>
48 <tr class="field"><th class="field-name">Updates:</th><td class="field-body">44</td>
49 </tr>
50 </tbody>
51 </table>
52 <hr />
53 <div class="contents topic" id="contents">
54 <p class="topic-title first">Contents</p>
55 <ul class="simple">
56 <li><a class="reference internal" href="#abstract" id="id1">Abstract</a></li>
57 <li><a class="reference internal" href="#motivation" id="id2">Motivation</a></li>
58 <li><a class="reference internal" href="#specification" id="id3">Specification</a><ul>
59 <li><a class="reference internal" href="#the-bad-news" id="id4">The bad news</a></li>
60 <li><a class="reference internal" href="#how-fast-can-md5-be-broken" id="id5">How fast can MD5 be broken?</a></li>
61 <li><a class="reference internal" href="#the-good-news" id="id6">The good news</a></li>
62 <li><a class="reference internal" href="#what-should-be-done" id="id7">What should be done</a><ul>
63 <li><a class="reference internal" href="#checksum-depreciation" id="id8">Checksum depreciation</a></li>
64 </ul>
65 </li>
66 </ul>
67 </li>
68 <li><a class="reference internal" href="#backwards-compatibility" id="id9">Backwards Compatibility</a></li>
69 <li><a class="reference internal" href="#references" id="id10">References</a></li>
70 <li><a class="reference internal" href="#thanks-to" id="id11">Thanks to</a></li>
71 <li><a class="reference internal" href="#copyright" id="id12">Copyright</a></li>
72 </ul>
73 </div>
74 <div class="section" id="abstract">
75 <h1><a class="toc-backref" href="#id1">Abstract</a></h1>
76 <p>While Manifest2 format allows multiple hashes, the question of which
77 checksums should be present, why, and the security implications of such
78 have never been resolved. This GLEP covers all of these issues, and
79 makes recommendations as to how to handle checksums both now, and in
80 future.</p>
81 </div>
82 <div class="section" id="motivation">
83 <h1><a class="toc-backref" href="#id2">Motivation</a></h1>
84 <p>This GLEP is being written as part of the work on signing the Portage
85 tree, but is only tangentially related to the actual signing of
86 Manifests. Checksums present one possible weak point in the overall
87 security of the tree - and a comprehensive security plan is needed.</p>
88 </div>
89 <div class="section" id="specification">
90 <h1><a class="toc-backref" href="#id3">Specification</a></h1>
91 <div class="section" id="the-bad-news">
92 <h2><a class="toc-backref" href="#id4">The bad news</a></h2>
93 <p>First of all, I'd like to cover the bad news in checksum security.
94 A much discussed point, as been the simple question: What is the
95 security of multiple independent checksums on the same data?
96 The most common position (and indeed the one previously held by myself),
97 is that multiple checksums would be an increase in security, but we
98 could not provably quantify the amount of security this added.
99 The really bad news, is that this position is completely and utterly
100 wrong. Many of you will be aghast at this. There is extremely little
101 added security in multiple checksums [J04]. For any set of checksums,
102 the actual strength lies in that of the strongest checksum.</p>
103 </div>
104 <div class="section" id="how-fast-can-md5-be-broken">
105 <h2><a class="toc-backref" href="#id5">How fast can MD5 be broken?</a></h2>
106 <p>For a general collision, not a pre-image attack, since the original
107 announcement by Wang et al [W04], the time required to break MD5 has
108 been massively reduced. Originally at 1 hour on a near-supercomputer
109 (IBM P690) and estimated at 64 hours with a Pentium-3 1.7Ghz. This has
110 gone down to less than in two years, to 17 seconds [K06a]!</p>
111 <p>08/2004 - 1 hour, IBM pSeries 690 (32x 1.7Ghz POWER4+) = 54.4 GHz-Hours
112 03/2005 - 8 hours, Pentium-M 1.6Ghz = 12.8 Ghz-Hours
113 11/2005 - 5 hours, Pentium-4 1.7Ghz = 8.5 Ghz-Hours
114 03/2006 - 1 minute, Pentium-4 3.2Ghz = .05 Ghz-Hours
115 04/2006 - 17 seconds, Pentium-4 3.2Ghz = .01 Ghz-Hours</p>
116 <p>If we accept a factor of 800x as a sample of how much faster a checksum
117 may be broken over the course of 2 years (MD5 using the above data is
118 &gt;2000x), then existing checksums do not stand a significant chance of
119 survival in the future. We should thus accept that whatever checksums we
120 are using today, will be broken in the near future, and plan as best as
121 possible. (A brief review [H04] of the present SHA1 attacks indicates an
122 improvement of ~600x in the same timespan).</p>
123 <p>And for those that claim implementation of these procedures is not yet
124 feasible, see [K06b] for an application that can produce two
125 self-extracting .exe files, with identical MD5s, and whatever payload
126 you want.</p>
127 </div>
128 <div class="section" id="the-good-news">
129 <h2><a class="toc-backref" href="#id6">The good news</a></h2>
130 <p>Of the checksums presently used by Manifest2, one stands close to being
131 completely broken: SHA1. The SHA2 series has suffered some attacks, but
132 still remains reasonably solid [G07],[K08]. No attacks against RIPEMD160
133 have been published, however it is constructed in the same manner as
134 MD5, SHA1 and SHA2, so is also vulnerable to the new methods of
135 cryptanalysis [H04].</p>
136 <p>To reduce the potential for future problems and any single checksum
137 break leading to a rapid decrease in security, we should incorporate the
138 strongest hash available from each family of checksums, and be prepared
139 to retire old checksums actively, unless there is a overriding reason to
140 keep a specific checksum.</p>
141 </div>
142 <div class="section" id="what-should-be-done">
143 <h2><a class="toc-backref" href="#id7">What should be done</a></h2>
144 <p>Portage should always try to verify all supported hashes that are
145 available in a Manifest2, starting with the strongest ones as maintained
146 by a preference list. Over time, the weaker checksums should be removed
147 from Manifest2 files, once all old Portage installations have had
148 sufficient time to upgrade. We should be prepared to add stronger
149 checksums wherever possible, and to remove those that have been
150 defeated.</p>
151 <p>An unsupported hash is not considered to be a failure unless no
152 supported hashes are available.</p>
153 <div class="section" id="checksum-depreciation">
154 <h3><a class="toc-backref" href="#id8">Checksum depreciation</a></h3>
155 <p>For the current Portage, SHA1 should be gradually removed, as presents
156 no advantages over SHA256. Beyond one specific problem (see the next
157 paragraph), we should add SHA512 (SHA2, 512 bit size), the Whirlpool
158 checksum (standardized checksum, with no known weaknesses). In future,
159 as stream-based checksums are developed (in response to the development
160 by NIST [AHS]), they should be considered and used.</p>
161 <p>There is one temporary stumbling block at hand - the existing Portage
162 infrastructure does not support SHA384/512 or Whirlpool, thus hampering
163 their immediate acceptance. SHA512 is available in Python 2.5, while
164 SHA1 is already available in Python 2.4. After Python2.5 is established
165 in a Gentoo media release, that would be a suitable time to remove SHA1
166 from Manifest2 files.</p>
167 </div>
168 </div>
169 </div>
170 <div class="section" id="backwards-compatibility">
171 <h1><a class="toc-backref" href="#id9">Backwards Compatibility</a></h1>
172 <p>Old versions of Portage may support and expect only specific checksums.
173 This is accounted for in the checksum depreciation discussion.</p>
174 </div>
175 <div class="section" id="references">
176 <h1><a class="toc-backref" href="#id10">References</a></h1>
177 <dl class="docutils">
178 <dt>[AHS] NIST (2007). &quot;NIST's Plan for New Cryptographic Hash Functions&quot;,</dt>
179 <dd>(Advanced Hash Standard). <a class="reference external" href="http://csrc.nist.gov/pki/HashWorkshop/">http://csrc.nist.gov/pki/HashWorkshop/</a></dd>
180 <dt>[BOBO06] Boneh, D. and Boyen, X. (2006). &quot;On the Impossibility of</dt>
181 <dd>Efficiently Combining Collision Resistant Hash Functions&quot;; Proceedings
182 of CRYPTO 2006, Dwork, C. (Ed.); Lecture Notes in Computer Science
183 4117, pp. 570-583. Available online from:
184 <a class="reference external" href="http://crypto.stanford.edu/~dabo/abstracts/hashing.html">http://crypto.stanford.edu/~dabo/abstracts/hashing.html</a></dd>
185 <dt>[H04] Hawkes, P. and Paddon, M. and Rose, G. (2004). &quot;On Corrective</dt>
186 <dd>Patterns for the SHA-2 Family&quot;. CRYPTO 2004 Cryptology ePrint Archive,
187 Report 2004/204. Available online from:
188 <a class="reference external" href="http://eprint.iacr.org/2004/207.pdf">http://eprint.iacr.org/2004/207.pdf</a></dd>
189 <dt>[J04] Joux, Antoie. (2004). &quot;Multicollisions in Iterated Hash</dt>
190 <dd>Functions - Application to Cascaded Constructions;&quot; Proceedings of
191 CRYPTO 2004, Franklin, M. (Ed); Lecture Notes in Computer Science
192 3152, pp. 306-316. Available online from:
193 <a class="reference external" href="http://web.cecs.pdx.edu/~teshrim/spring06/papers/general-attacks/multi-joux.pdf">http://web.cecs.pdx.edu/~teshrim/spring06/papers/general-attacks/multi-joux.pdf</a></dd>
194 <dt>[K06a] Klima, V. (2006). &quot;Tunnels in Hash Functions: MD5 Collisions</dt>
195 <dd>Within a Minute&quot;. Cryptology ePrint Archive, Report 2006/105.
196 Available online from: <a class="reference external" href="http://eprint.iacr.org/2006/105.pdf">http://eprint.iacr.org/2006/105.pdf</a></dd>
197 <dt>[K06b] Klima, V. (2006). &quot;Note and links to high-speed MD5 collision</dt>
198 <dd>proof of concept tools&quot;. Available online from:
199 <a class="reference external" href="http://cryptography.hyperlink.cz/2006/trick.txt">http://cryptography.hyperlink.cz/2006/trick.txt</a></dd>
200 <dt>[K08] Klima, V. (2008). &quot;On Collisions of Hash Functions Turbo SHA-2&quot;.</dt>
201 <dd>Cryptology ePrint Archive, Report 2008/003. Available online from:
202 <a class="reference external" href="http://eprint.iacr.org/2008/003.pdf">http://eprint.iacr.org/2008/003.pdf</a></dd>
203 <dt>[G07] Gligoroski, D. and Knapskog, S.J. (2007). &quot;Turbo SHA-2&quot;.</dt>
204 <dd>Cryptology ePrint Archive, Report 2007/403. Available online from:
205 <a class="reference external" href="http://eprint.iacr.org/2007/403.pdf">http://eprint.iacr.org/2007/403.pdf</a></dd>
206 <dt>[W04] Wang, X. et al: &quot;Collisions for Hash Functions MD4, MD5,</dt>
207 <dd>HAVAL-128 and RIPEMD&quot;, rump session, CRYPTO 2004, Cryptology ePrint
208 Archive, Report 2004/199, first version (August 16, 2004), second
209 version (August 17, 2004). Available online from:
210 <a class="reference external" href="http://eprint.iacr.org/2004/199.pdf">http://eprint.iacr.org/2004/199.pdf</a></dd>
211 </dl>
212 </div>
213 <div class="section" id="thanks-to">
214 <h1><a class="toc-backref" href="#id11">Thanks to</a></h1>
215 <dl class="docutils">
216 <dt>I'd like to thank the following folks, in no specific order:</dt>
217 <dd><ul class="first last simple">
218 <li>Ciaran McCreesh (ciaranm) - for pointing out the Joux (2004) paper,
219 and also being stubborn enough in not accepting a partial solution.</li>
220 <li>Marius Mauch (genone), Zac Medico (zmedico) and Brian Harring
221 (ferringb): for being knowledgeable about the Portage Manifest2
222 codebase.</li>
223 </ul>
224 </dd>
225 </dl>
226 </div>
227 <div class="section" id="copyright">
228 <h1><a class="toc-backref" href="#id12">Copyright</a></h1>
229 <p>Copyright (c) 2006 by Robin Hugh Johnson. This material may be
230 distributed only subject to the terms and conditions set forth in the
231 Open Publication License, v1.0.</p>
232 <p>vim: tw=72 ts=2 expandtab:</p>
233 </div>
235 </div>
236 <div class="footer">
237 <hr class="footer" />
238 <a class="reference external" href="glep-0059.txt">View document source</a>.
239 Generated on: 2008-10-22 18:02 UTC.
240 Generated by <a class="reference external" href="http://docutils.sourceforge.net/">Docutils</a> from <a class="reference external" href="http://docutils.sourceforge.net/rst.html">reStructuredText</a> source.
242 </div>
243 </body>
244 </html>

  ViewVC Help
Powered by ViewVC 1.1.20