Annotation of www/security.html, Revision 1.83
1.1 deraadt 1: <!DOCTYPE HTML PUBLIC "-//IETF//DTD HTML Strict//EN">
2: <html>
3: <head>
1.20 deraadt 4: <title>OpenBSD Security</title>
1.1 deraadt 5: <link rev=made href=mailto:www@openbsd.org>
6: <meta name="resource-type" content="document">
7: <meta name="description" content="OpenBSD advisories">
8: <meta name="keywords" content="openbsd,main">
9: <meta name="distribution" content="global">
1.45 deraadt 10: <meta name="copyright" content="This document copyright 1997,1998 by OpenBSD.">
1.1 deraadt 11: </head>
12:
13: <BODY BGCOLOR="#FFFFFF" TEXT="#000000" LINK="#23238E">
14:
1.77 deraadt 15: <img alt="[OpenBSD]" height=30 width=141 SRC="images/smalltitle.gif">
1.1 deraadt 16:
1.56 deraadt 17: <hr>
18: <a href=#21>For 2.1 security advisories, please refer here</a>.<br>
19: <a href=#22>For 2.2 security advisories, please refer here</a>.<br>
20: <a href=#23>For 2.3 security advisories, please refer here</a>.<br>
1.75 deraadt 21: <a href=#24>For 2.4 security advisories, please refer here</a>.<br>
1.56 deraadt 22: <hr>
23:
1.2 deraadt 24: <p>
1.12 deraadt 25: <h3><font color=#e00000><strong>OpenBSD Security Views</strong></font></h3>
1.22 deraadt 26:
1.14 deraadt 27: OpenBSD believes in strong security. Our aspiration is to be NUMBER
1.22 deraadt 28: ONE in the industry for security (if we are not already there). Our
29: open software development model permits us to take a more
30: uncompromising view towards increased security than Sun, SGI, IBM, HP,
31: or other vendors are able to. We can make changes the vendors would
1.27 deraadt 32: not make. Also, since OpenBSD is exported with <a href=crypto.html>
1.45 deraadt 33: cryptography</a>, we are able to take cryptographic approaches towards
34: fixing security problems.<p>
1.18 deraadt 35:
1.45 deraadt 36: Like many readers of the
1.13 deraadt 37: <a href=http://www.geek-girl.com/bugtraq/index.html>
1.18 deraadt 38: BUGTRAQ mailing list</a>,
1.45 deraadt 39: we believe in full disclosure of security problems. Security
40: information moves very fast in cracker circles. On the other hand,
41: our experience is that coding and releasing of proper security fixes
42: typically requires about an hour of work -- very fast fix turnaround
43: is possible. Thus we think that full disclosure helps the people who
1.22 deraadt 44: really care about security.<p>
1.15 deraadt 45:
1.12 deraadt 46: Our security auditing team typically has between six and twelve
1.45 deraadt 47: members who continue to search for and fix new security holes. We
48: have been auditing since the summer of 1996. The process we follow to
49: increase security is simply a comprehensive file-by-file analysis of
50: every critical software component. Flaws have been found in just
51: about every area of the system. Entire new classes of security
1.80 espie 52: problems have been found during our audit, and often source code
1.45 deraadt 53: which had been audited earlier needs re-auditing with these new flaws
54: in mind. Code often gets audited multiple times, and by multiple
55: people with different auditing skills.<p>
1.12 deraadt 56:
1.31 deraadt 57: Some members of our security auditing team work for
58: <a href=http://www.secnet.com>Secure Networks</a>, the company that
1.32 deraadt 59: makes the industry's premier network security scanning software
60: package Ballista.
1.31 deraadt 61: This company does a lot of security research, and this fits in well
1.45 deraadt 62: with the OpenBSD stance. OpenBSD passes Ballista's tests with flying
63: colours.<p>
1.31 deraadt 64:
1.34 deraadt 65: Another facet of our security auditing process is its proactiveness.
1.45 deraadt 66: In most cases we have found that the determination of exploitability
67: is not an issue. During our ongoing auditing process we find many
68: bugs, and endeavor to fix them even though exploitability is not
69: proven. We fix the bug, and we move on to find other bugs to fix. We
70: have fixed many simple and obvious careless programming errors in code
71: and only months later discovered that the problems were in fact
72: exploitable. (Or, more likely someone on
73: <a href=http://www.geek-girl.com/bugtraq/index.html>BUGTRAQ</a>
74: would report that other operating systems were vulnerable to a `newly
75: discovered problem', and then it would be discovered that OpenBSD had
76: been fixed in a previous release). In other cases we have been saved
77: from full exploitability of complex step-by-step attacks because we
78: had fixed one of the intermediate steps. An example of where we
79: managed such a success is the
1.30 deraadt 80: <a href=http://www.secnet.com/sni-advisories/sni-19.bsd.lpd.advisory.html>
1.35 deraadt 81: lpd advisory from Secure Networks.</a><p>
1.29 deraadt 82:
1.45 deraadt 83: Our proactive auditing process has really paid off. Statements like
1.35 deraadt 84: ``This problem was fixed in OpenBSD about 6 months ago'' have become
1.45 deraadt 85: commonplace in security forums like
86: <a href=http://www.geek-girl.com/bugtraq/index.html>BUGTRAQ</a>.<p>
1.35 deraadt 87:
1.45 deraadt 88: The most intense part of our security auditing happened immediately
1.80 espie 89: before the OpenBSD 2.0 release and during the 2.0->2.1 transition,
1.45 deraadt 90: over the last third of 1996 and first half of 1997. Thousands (yes,
91: thousands) of security issues were fixed rapidly over this year-long
92: period; bugs like the standard buffer overflows, protocol
93: implementation weaknesses, information gathering, and filesystem
94: races. Hence most of the security problems that we encountered were
95: fixed before our 2.1 release, and then a far smaller number needed
96: fixing for our 2.2 release. We do not find as many problems anymore,
97: it is simply a case of diminishing returns. Recently the security
98: problems we find and fix tend to be significantly more obscure or
99: complicated. Still we will persist for a number of reasons:<p>
1.36 deraadt 100:
1.35 deraadt 101: <ul>
1.45 deraadt 102: <li>Occasionally we find a simple problem we missed earlier. Doh!
1.35 deraadt 103: <li>Security is like an arms race; the best attackers will continue
1.45 deraadt 104: to search for more complicated exploits, so we will too.
105: <li>Finding and fixing subtle flaws in complicated software is
106: a lot of fun.
1.35 deraadt 107: </ul>
1.15 deraadt 108:
1.14 deraadt 109: The auditing process is not over yet, and as you can see we continue
1.28 deraadt 110: to find and fix new security flaws.<p>
1.12 deraadt 111:
1.75 deraadt 112: <a name=24></a>
113: <p>
114: <h3><font color=#e00000><strong>OpenBSD 2.4 Security Advisories</strong></font></h3>
115: These are the OpenBSD 2.4 advisories -- all these problems are solved
116: in <a href=anoncvs.html>OpenBSD current</a>. Obviously, all the
117: OpenBSD 2.3 advisories listed below are fixed in OpenBSD 2.4.
118:
119: <ul>
1.83 ! deraadt 120: <li><a href=errata.html#rst>Feb 11, 1999: TCP/IP RST handling was sloppy.
! 121: (patch included).</a>
1.81 deraadt 122: <li><a href=errata.html#bootpd>Nov 27, 1998: There is a remotely exploitable
123: problem in bootpd(8). (patch included).</a>
1.82 deraadt 124: <li><a href=errata.html#termcap>Nov 19, 1998: There is a possibly locally
125: exploitable problem relating to environment variables in termcap
126: and curses. (patch included).</a>
1.78 deraadt 127: <li><a href=errata.html#tcpfix>Nov 13, 1998: There is a remote machine lockup
128: bug in the TCP decoding kernel. (patch included).</a>
1.75 deraadt 129: </ul>
130:
1.58 deraadt 131: <a name=23></a>
1.12 deraadt 132: <p>
1.52 deraadt 133: <h3><font color=#e00000><strong>OpenBSD 2.3 Security Advisories</strong></font></h3>
1.73 deraadt 134: These are the OpenBSD 2.3 advisories -- all these problems are solved
135: in <a href=anoncvs.html>OpenBSD current</a>. Obviously, all the
136: OpenBSD 2.2 advisories listed below are fixed in OpenBSD 2.3.
1.53 matthieu 137:
138: <ul>
1.81 deraadt 139: <li><a href=errata23.html#bootpd>Nov 27, 1998: There is a remotely exploitable
140: problem in bootpd(8). (patch included).</a>
1.78 deraadt 141: <li><a href=errata23.html#tcpfix>Nov 13, 1998: There is a remote machine lockup
142: bug in the TCP decoding kernel. (patch included).</a>
1.76 aaron 143: <li><a href=errata23.html#fdalloc>Jul 2, 1998: setuid and setgid processes
1.72 deraadt 144: should not be executed with fd slots 0, 1, or 2 free.
145: (patch included).</a>
1.79 deraadt 146: <li><a href=errata23.html#resolver>August 31, 1998: A benign looking resolver buffer overflow bug was re-introduced accidentally (patches included).</a>
1.76 aaron 147: <li><a href=errata23.html#xlib>June 6, 1998: Further problems with the X
1.71 deraadt 148: libraries (patches included).</a>
1.76 aaron 149: <li><a href=errata23.html#pctr>June 4, 1998: on non-Intel i386 machines, any user
1.72 deraadt 150: can use pctr(4) to crash the machine.</a>
1.76 aaron 151: <li><a href=errata23.html#kill>May 17, 1998: kill(2) of setuid/setgid target
1.66 deraadt 152: processes too permissive (4th revision patch included).</a>
1.76 aaron 153: <li><a href=errata23.html#immutable>May 11, 1998: mmap() permits partial bypassing
1.60 deraadt 154: of immutable and append-only file flags. (patch included).</a>
1.76 aaron 155: <li><a href=errata23.html#xterm-xaw>May 1, 1998: Buffer overflow in xterm and Xaw
1.58 deraadt 156: (CERT advisory VB-98.04) (patch included).</a>
1.76 aaron 157: <li><a href=errata23.html#ipsec>May 5, 1998: Incorrect handling of IPSEC packets
1.59 deraadt 158: if IPSEC is enabled (patch included).</a>
1.53 matthieu 159: </ul>
1.9 deraadt 160:
1.58 deraadt 161: <a name=22></a>
1.9 deraadt 162: <p>
1.12 deraadt 163: <h3><font color=#e00000><strong>OpenBSD 2.2 Security Advisories</strong></font></h3>
1.45 deraadt 164: These are the OpenBSD 2.2 advisories. All these problems are solved
1.55 deraadt 165: in <a href=23.html>OpenBSD 2.3</a>. Some of these problems
1.45 deraadt 166: still exist in other operating systems. (The supplied patches are for
167: OpenBSD 2.2; they may or may not work on OpenBSD 2.1).
1.9 deraadt 168:
169: <ul>
1.72 deraadt 170: <li><a href=errata22.html#ipsec>May 5, 1998: Incorrect handling of IPSEC
171: packets if IPSEC is enabled (patch included).</a>
172: <li><a href=errata22.html#xterm-xaw>May 1, 1998: Buffer overflow in xterm
173: and Xaw (CERT advisory VB-98.04) (patch included).</a>
174: <li><a href=errata22.html#uucpd>Apr 22, 1998: Buffer overflow in uucpd
175: (patch included).</a>
176: <li><a href=errata22.html#rmjob>Apr 22, 1998: Buffer mismanagement in lprm
177: (patch included).</a>
178: <li><a href=errata22.html#ping>Mar 31, 1998: Overflow in ping -R (patch included).</a>
179: <li><a href=errata22.html#named>Mar 30, 1998: Overflow in named fake-iquery
1.59 deraadt 180: (patch included).</a>
1.72 deraadt 181: <li><a href=errata22.html#mountd>Mar 2, 1998: Accidental NFS filesystem
182: export (patch included).</a>
183: <li><a href=advisories/mmap>Feb 26, 1998: Read-write mmap() flaw.</a>
184: Revision 3 of the patch is available <a href=errata22.html#mmap>here</a>
1.59 deraadt 185: <li><a href=advisories/sourceroute>Feb 19, 1998: Sourcerouted Packet
186: Acceptance.</a>
1.50 deraadt 187: A patch is available <a href=errata22.html#sourceroute>here</a>.
1.72 deraadt 188: <li><a href=errata22.html#ruserok>Feb 13, 1998: Setuid coredump & Ruserok()
189: flaw (patch included).</a>
190: <li><a href=errata22.html#ldso>Feb 9, 1998: MIPS ld.so flaw (patch included).</a>
191: <li><a href=errata22.html#f00f>Dec 10, 1997: Intel P5 f00f lockup
1.59 deraadt 192: (patch included).</a>
1.1 deraadt 193: </ul>
194:
1.58 deraadt 195: <a name=21></a>
1.21 deraadt 196: <p>
1.52 deraadt 197: <h3><font color=#e00000><strong>OpenBSD 2.1 Security Advisories</strong></font></h3>
198: These are the OpenBSD 2.1 advisories. All these problems are solved
199: in <a href=22.html>OpenBSD 2.2</a>. Some of these problems still
200: exist in other operating systems. (If you are running OpenBSD 2.1, we
201: would strongly recommend an upgrade to the newest release, as this
202: patch list only attempts at fixing the most important security
203: problems. In particular, OpenBSD 2.2 fixes numerous localhost
204: security problems. Many of those problems were solved in ways which
205: make it hard for us to provide patches).
206:
207: <ul>
1.72 deraadt 208: <li><a href=advisories/signals>Sep 15, 1997: Deviant Signals (patch included)</a>
1.59 deraadt 209: <li><a href=advisories/rfork>Aug 2, 1997: Rfork() system call flaw
210: (patch included)</a>
211: <li><a href=advisories/procfs>Jun 24, 1997: Procfs flaws (patch included)</a>
1.52 deraadt 212: </ul>
1.51 deraadt 213:
214: <p>
1.21 deraadt 215: <h3><font color=#e00000><strong>Watching our Security Changes</strong></font></h3>
216: Since we take a proactive stance with security, we are continually
217: finding and fixing new security problems. Not all of these problems
1.80 espie 218: get widely reported because (as stated earlier) many of them are not
1.45 deraadt 219: confirmed to be exploitable; many simple bugs we fix do turn out to
220: have security consequences we could not predict. We do not have the
221: time resources to make these changes available in the above format.<p>
1.21 deraadt 222:
223: Thus there are usually minor security fixes in the current source code
224: beyond the previous major OpenBSD release. We make a limited
1.45 deraadt 225: guarantee that these problems are of minimal impact and unproven
1.44 ian 226: exploitability. If we discover that a problem definitely matters for
1.45 deraadt 227: security, patches will show up here <strong>VERY</strong> quickly.<p>
1.21 deraadt 228:
1.45 deraadt 229: People who are really concerned with security can do a number of
230: things:<p>
1.21 deraadt 231:
232: <ul>
233: <li>If you understand security issues, watch our
1.27 deraadt 234: <a href=mail.html>source-changes mailing list</a> and keep an
1.23 deraadt 235: eye out for things which appear security related. Since
1.21 deraadt 236: exploitability is not proven for many of the fixes we make,
237: do not expect the relevant commit message to say "SECURITY FIX!".
238: If a problem is proven and serious, a patch will be available
239: here very shortly after.
240: <li>Track our current source code tree, and teach yourself how to do a
1.29 deraadt 241: complete system build from time to time (read /usr/src/Makefile
242: carefully). Users can make the assumption that the current
243: source tree always has stronger security than the previous release.
1.45 deraadt 244: However, building your own system from source code is not trivial;
245: it is nearly 300MB of source code, and problems do occur as we
246: transition between major releases.
1.29 deraadt 247: <li>Install a binary <a href=snapshots.html>snapshot</a> for your
1.80 espie 248: architecture, which are made available fairly often. For
1.29 deraadt 249: instance, an i386 snapshot is typically made available weekly.
1.21 deraadt 250: </ul>
251:
1.9 deraadt 252: <p>
1.12 deraadt 253: <h3><font color=#e00000><strong>Other Resources</strong></font></h3>
1.3 deraadt 254: Other security advisories that have (in the past) affected OpenBSD can
1.4 deraadt 255: be found at the <a href=http://www.secnet.com/nav1.html>Secure Networks archive</a>.
1.25 deraadt 256: Some OpenBSD audit team members worked with Secure Networks on discovering
257: and solving the problems detailed in some of their security advisories.
1.3 deraadt 258:
1.5 deraadt 259: <p> If you find a new security problem, you can mail it to
1.6 deraadt 260: <a href=mailto:deraadt@openbsd.org>deraadt@openbsd.org</a>.
1.7 deraadt 261: <br>
1.5 deraadt 262: If you wish to PGP encode it (but please only do so if privacy is very
1.27 deraadt 263: urgent, since it is inconvenient) use this <a href=advisories/pgpkey>pgp key</a>.
1.5 deraadt 264:
1.2 deraadt 265: <hr>
1.68 pauls 266: <a href=index.html><img height=24 width=24 src=back.gif border=0 alt=OpenBSD></a>
1.24 deraadt 267: <a href=mailto:www@openbsd.org>www@openbsd.org</a>
268: <br>
1.83 ! deraadt 269: <small>$OpenBSD: security.html,v 1.82 1999/02/09 09:00:22 deraadt Exp $</small>
1.1 deraadt 270:
1.24 deraadt 271: </body>
272: </html>