/[gentoo]/xml/htdocs/proj/en/glep/glep-0059.html
Gentoo

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

Parent Directory Parent Directory | Revision Log Revision Log | View Patch Patch

Revision 1.3 Revision 1.6
… …
2<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"> 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"> 3<html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en" lang="en">
4 4
5<head> 5<head>
6 <meta http-equiv="Content-Type" content="text/html; charset=utf-8" /> 6 <meta http-equiv="Content-Type" content="text/html; charset=utf-8" />
7 <meta name="generator" content="Docutils 0.5: http://docutils.sourceforge.net/" /> 7 <meta name="generator" content="Docutils 0.6: http://docutils.sourceforge.net/" />
8 <title>GLEP 59 -- Manifest2 hash policies and security implications</title> 8 <title>GLEP 59 -- Manifest2 hash policies and security implications</title>
9 <link rel="stylesheet" href="tools/glep.css" type="text/css" /></head> 9 <link rel="stylesheet" href="tools/glep.css" type="text/css" /></head>
10<body bgcolor="white"> 10<body bgcolor="white">
11<table class="navigation" cellpadding="0" cellspacing="0" 11<table class="navigation" cellpadding="0" cellspacing="0"
12 width="100%" border="0"> 12 width="100%" border="0">
… …
25<tbody valign="top"> 25<tbody valign="top">
26<tr class="field"><th class="field-name">GLEP:</th><td class="field-body">59</td> 26<tr class="field"><th class="field-name">GLEP:</th><td class="field-body">59</td>
27</tr> 27</tr>
28<tr class="field"><th class="field-name">Title:</th><td class="field-body">Manifest2 hash policies and security implications</td> 28<tr class="field"><th class="field-name">Title:</th><td class="field-body">Manifest2 hash policies and security implications</td>
29</tr> 29</tr>
30<tr class="field"><th class="field-name">Version:</th><td class="field-body">1.3</td> 30<tr class="field"><th class="field-name">Version:</th><td class="field-body">1.5</td>
31</tr> 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">2008/10/28 07:45:44</a></td> 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">2010/01/31 07:55:45</a></td>
33</tr> 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> 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> 35</tr>
36<tr class="field"><th class="field-name">Status:</th><td class="field-body">Draft</td> 36<tr class="field"><th class="field-name">Status:</th><td class="field-body">Draft</td>
37</tr> 37</tr>
… …
41</tr> 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> 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> 43</tr>
44<tr class="field"><th class="field-name">Created:</th><td class="field-body">October 2006</td> 44<tr class="field"><th class="field-name">Created:</th><td class="field-body">October 2006</td>
45</tr> 45</tr>
46<tr class="field"><th class="field-name">Updated:</th><td class="field-body">November 2007, June 2008, July 2008, October 2008</td> 46<tr class="field"><th class="field-name">Updated:</th><td class="field-body">November 2007, June 2008, July 2008, October 2008, January 2010</td>
47</tr> 47</tr>
48<tr class="field"><th class="field-name">Updates:</th><td class="field-body">44</td> 48<tr class="field"><th class="field-name">Updates:</th><td class="field-body">44</td>
49</tr> 49</tr>
50<tr class="field"><th class="field-name">Post-History:</th><td class="field-body"></td> 50<tr class="field"><th class="field-name">Post-History:</th><td class="field-body">December 2009, January 2010</td>
51</tr> 51</tr>
52</tbody> 52</tbody>
53</table> 53</table>
54<hr /> 54<hr />
55<div class="contents topic" id="contents"> 55<div class="contents topic" id="contents">
… …
59<li><a class="reference internal" href="#motivation" id="id2">Motivation</a></li> 59<li><a class="reference internal" href="#motivation" id="id2">Motivation</a></li>
60<li><a class="reference internal" href="#specification" id="id3">Specification</a><ul> 60<li><a class="reference internal" href="#specification" id="id3">Specification</a><ul>
61<li><a class="reference internal" href="#the-bad-news" id="id4">The bad news</a></li> 61<li><a class="reference internal" href="#the-bad-news" id="id4">The bad news</a></li>
62<li><a class="reference internal" href="#how-fast-can-md5-be-broken" id="id5">How fast can MD5 be broken?</a></li> 62<li><a class="reference internal" href="#how-fast-can-md5-be-broken" id="id5">How fast can MD5 be broken?</a></li>
63<li><a class="reference internal" href="#the-good-news" id="id6">The good news</a></li> 63<li><a class="reference internal" href="#the-good-news" id="id6">The good news</a></li>
64<li><a class="reference internal" href="#what-should-be-done" id="id7">What should be done</a><ul> 64<li><a class="reference internal" href="#what-should-be-done" id="id7">What should be done</a></li>
65<li><a class="reference internal" href="#checksum-depreciation" id="id8">Checksum depreciation</a></li> 65<li><a class="reference internal" href="#checksum-depreciation-timing" id="id8">Checksum depreciation timing</a></li>
66</ul>
67</li>
68</ul> 66</ul>
69</li> 67</li>
70<li><a class="reference internal" href="#backwards-compatibility" id="id9">Backwards Compatibility</a></li> 68<li><a class="reference internal" href="#backwards-compatibility" id="id9">Backwards Compatibility</a></li>
71<li><a class="reference internal" href="#references" id="id10">References</a></li> 69<li><a class="reference internal" href="#references" id="id10">References</a></li>
72<li><a class="reference internal" href="#thanks-to" id="id11">Thanks to</a></li> 70<li><a class="reference internal" href="#thanks-to" id="id11">Thanks to</a></li>
… …
98The most common position (and indeed the one previously held by myself), 96The most common position (and indeed the one previously held by myself),
99is that multiple checksums would be an increase in security, but we 97is that multiple checksums would be an increase in security, but we
100could not provably quantify the amount of security this added. 98could not provably quantify the amount of security this added.
101The really bad news, is that this position is completely and utterly 99The really bad news, is that this position is completely and utterly
102wrong. Many of you will be aghast at this. There is extremely little 100wrong. Many of you will be aghast at this. There is extremely little
103added security in multiple checksums [J04]. For any set of checksums, 101added security in multiple checksums as noted by Joux [J04]. For any set
104the actual strength lies in that of the strongest checksum.</p> 102of checksums, the actual strength lies in that of the strongest
103checksum.</p>
104<p>Wang et al [W04] extended Joux's [J04] work on SHA-0 to cover MD4, MD5,
105HAVAL-128 and RIPEMD families of hashes.</p>
105</div> 106</div>
106<div class="section" id="how-fast-can-md5-be-broken"> 107<div class="section" id="how-fast-can-md5-be-broken">
107<h2><a class="toc-backref" href="#id5">How fast can MD5 be broken?</a></h2> 108<h2><a class="toc-backref" href="#id5">How fast can MD5 be broken?</a></h2>
108<p>For a general collision, not a pre-image attack, since the original 109<p>For a general collision, not a pre-image attack, since the announcement
109announcement by Wang et al [W04], the time required to break MD5 has 110by Wang et al [W04], the time required to break MD5 has been massively
110been massively reduced. Originally at 1 hour on a near-supercomputer 111reduced. Originally at 1 hour on a near-supercomputer (IBM P690) and
111(IBM P690) and estimated at 64 hours with a Pentium-3 1.7Ghz. This has 112estimated at 64 hours with a Pentium-3 1.7Ghz. This has gone down to
112gone down to less than in two years, to 17 seconds [K06a]!</p> 113less than in two years, to 17 seconds [K06a].</p>
113<p>08/2004 - 1 hour, IBM pSeries 690 (32x 1.7Ghz POWER4+) = 54.4 GHz-Hours 114<p>08/2004 - 1 hour, IBM pSeries 690 (32x 1.7Ghz POWER4+) = 54.4 GHz-Hours
11403/2005 - 8 hours, Pentium-M 1.6Ghz = 12.8 Ghz-Hours 11503/2005 - 8 hours, Pentium-M 1.6Ghz = 12.8 Ghz-Hours
11511/2005 - 5 hours, Pentium-4 1.7Ghz = 8.5 Ghz-Hours 11611/2005 - 5 hours, Pentium-4 1.7Ghz = 8.5 Ghz-Hours
11603/2006 - 1 minute, Pentium-4 3.2Ghz = .05 Ghz-Hours 11703/2006 - 1 minute, Pentium-4 3.2Ghz = .05 Ghz-Hours
11704/2006 - 17 seconds, Pentium-4 3.2Ghz = .01 Ghz-Hours</p> 11804/2006 - 17 seconds, Pentium-4 3.2Ghz = .01 Ghz-Hours</p>
118<p>If we accept a factor of 800x as a sample of how much faster a checksum 119<p>If we accept a factor of 800x as a sample of how much faster a checksum
119may be broken over the course of 2 years (MD5 using the above data is 120may be broken over the course of 2 years (MD5 using the above data is
120&gt;2000x), then existing checksums do not stand a significant chance of 121&gt;2000x), then existing checksums do not stand a significant chance of
121survival in the future. We should thus accept that whatever checksums we 122survival in the future. We should thus accept that whatever checksums we
122are using today, will be broken in the near future, and plan as best as 123are using today, will be broken in the near future, and plan as best as
123possible. (A brief review [H04] of the present SHA1 attacks indicates an 124possible. (A brief review [H04] of the SHA1 attacks indicates an
124improvement of ~600x in the same timespan).</p> 125improvement of ~600x in the same timespan).</p>
125<p>And for those that claim implementation of these procedures is not yet 126<p>And for those that claim implementation of these procedures is not yet
126feasible, see [K06b] for an application that can produce two 127feasible, see [K06b] for an application that can produce two
127self-extracting .exe files, with identical MD5s, and whatever payload 128self-extracting EXE files, with identical MD5s, and whatever payload you
128you want.</p> 129want.</p>
129</div> 130</div>
130<div class="section" id="the-good-news"> 131<div class="section" id="the-good-news">
131<h2><a class="toc-backref" href="#id6">The good news</a></h2> 132<h2><a class="toc-backref" href="#id6">The good news</a></h2>
132<p>Of the checksums presently used by Manifest2, one stands close to being 133<p>Of the checksums presently used by Manifest2 (SHA1, SHA256, RIPEMD160),
133completely broken: SHA1. The SHA2 series has suffered some attacks, but 134one stands close to being completely broken: SHA1; and another is
134still remains reasonably solid [G07],[K08]. No attacks against RIPEMD160 135significantly weakened: RIPEMD160. The SHA2 series has suffered some
135have been published, however it is constructed in the same manner as 136attacks, but still remains reasonably solid [G07],[K08].</p>
136MD5, SHA1 and SHA2, so is also vulnerable to the new methods of
137cryptanalysis [H04].</p>
138<p>To reduce the potential for future problems and any single checksum 137<p>To reduce the potential for future problems and any single checksum
139break leading to a rapid decrease in security, we should incorporate the 138break leading to a rapid decrease in security, we should incorporate the
140strongest hash available from each family of checksums, and be prepared 139strongest hash available from each family of checksums, and be prepared
141to retire old checksums actively, unless there is a overriding reason to 140to retire old checksums actively, unless there is a overriding reason to
142keep a specific checksum.</p> 141keep a specific checksum, such as part of a migration plan.</p>
143</div> 142</div>
144<div class="section" id="what-should-be-done"> 143<div class="section" id="what-should-be-done">
145<h2><a class="toc-backref" href="#id7">What should be done</a></h2> 144<h2><a class="toc-backref" href="#id7">What should be done</a></h2>
146<p>Portage should always try to verify all supported hashes that are 145<p>Portage should always try to verify all supported hashes that are
147available in a Manifest2, starting with the strongest ones as maintained 146available in a Manifest2, starting with the strongest ones as maintained
148by a preference list. Over time, the weaker checksums should be removed 147by a preference list. Over time, the weaker checksums should be removed
149from Manifest2 files, once all old Portage installations have had 148from Manifest2 files, once all old Portage installations have had
150sufficient time to upgrade. We should be prepared to add stronger 149sufficient time to upgrade. We should be prepared to add stronger
151checksums wherever possible, and to remove those that have been 150checksums wherever possible, and to remove those that have been
152defeated.</p> 151defeated.</p>
152<p>As soon as feasible, we should add the SHA512 and WHIRLPOOL algorithms.
153In future, as stream-based checksums are developed (in response to the
154development by NIST [AHS]), they should be considered and used.</p>
155<p>The SHA512 algorithm is available in Python 2.5, which has been a
156dependency of Portage since approximately Python 2.1.6.13.</p>
157<p>The WHIRLPOOL checksum is not available within the PyCrypto library or
158hashlib that is part of Python 2.5, but there are multiple alternative
159Python implementations available, ranging from pure Python to C-based
160(python-mhash).</p>
153<p>An unsupported hash is not considered to be a failure unless no 161<p>The existence unsupported hash is not considered to be a failure unless
154supported hashes are available.</p> 162no supported hashes are available for a given Manifest entry.</p>
163</div>
155<div class="section" id="checksum-depreciation"> 164<div class="section" id="checksum-depreciation-timing">
156<h3><a class="toc-backref" href="#id8">Checksum depreciation</a></h3> 165<h2><a class="toc-backref" href="#id8">Checksum depreciation timing</a></h2>
157<p>For the current Portage, SHA1 should be gradually removed, as presents 166<p>For the current Portage, both SHA1 and RIPEMD160 should be immediately
158no advantages over SHA256. Beyond one specific problem (see the next 167removed, as they present no advantages over the already present SHA256.
159paragraph), we should add SHA512 (SHA2, 512 bit size), the Whirlpool 168SHA256 cannot be replaced immediately with SHA512, as existing Portage
160checksum (standardized checksum, with no known weaknesses). In future, 169versions need at least one supported algorithm present (SHA256 support
161as stream-based checksums are developed (in response to the development 170was added in June 2006), so it must be retained for some while.</p>
162by NIST [AHS]), they should be considered and used.</p> 171<p>Immediately:
163<p>There is one temporary stumbling block at hand - the existing Portage 172- Add WHIRLPOOL and SHA512.
164infrastructure does not support SHA384/512 or Whirlpool, thus hampering 173- Remove SHA1 and RIPEMD160.</p>
165their immediate acceptance. SHA512 is available in Python 2.5, while 174<p>After the majority of Portage installations include SHA512 support:
166SHA1 is already available in Python 2.4. After Python2.5 is established 175- Remove SHA256.</p>
167in a Gentoo media release, that would be a suitable time to remove SHA1
168from Manifest2 files.</p>
169</div>
170</div> 176</div>
171</div> 177</div>
172<div class="section" id="backwards-compatibility"> 178<div class="section" id="backwards-compatibility">
173<h1><a class="toc-backref" href="#id9">Backwards Compatibility</a></h1> 179<h1><a class="toc-backref" href="#id9">Backwards Compatibility</a></h1>
174<p>Old versions of Portage may support and expect only specific checksums. 180<p>Old versions of Portage may support and expect only specific checksums.
… …
226</dd> 232</dd>
227</dl> 233</dl>
228</div> 234</div>
229<div class="section" id="copyright"> 235<div class="section" id="copyright">
230<h1><a class="toc-backref" href="#id12">Copyright</a></h1> 236<h1><a class="toc-backref" href="#id12">Copyright</a></h1>
231<p>Copyright (c) 2006 by Robin Hugh Johnson. This material may be 237<p>Copyright (c) 2006-2010 by Robin Hugh Johnson. This material may be
232distributed only subject to the terms and conditions set forth in the 238distributed only subject to the terms and conditions set forth in the
233Open Publication License, v1.0.</p> 239Open Publication License, v1.0.</p>
234<p>vim: tw=72 ts=2 expandtab:</p> 240<p>vim: tw=72 ts=2 expandtab:</p>
235</div> 241</div>
236 242
237</div> 243</div>
238<div class="footer"> 244<div class="footer">
239<hr class="footer" /> 245<hr class="footer" />
240<a class="reference external" href="glep-0059.txt">View document source</a>. 246<a class="reference external" href="glep-0059.txt">View document source</a>.
241Generated on: 2008-10-28 07:47 UTC. 247Generated on: 2010-01-31 07:55 UTC.
242Generated 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. 248Generated 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.
243 249
244</div> 250</div>
245</body> 251</body>
246</html> 252</html>
Legend:
Removed from v.1.3  
changed lines
  Added in v.1.6
  ViewVC Help
Powered by ViewVC 1.1.13