I think all in all That it should be considered!<br><br>On 5/1/07, Steven Adair &lt;<a href="mailto:steven@securityzone.org">steven@securityzone.org</a>&gt; wrote:<br>&gt; I think a good share of the time when someone states that the DoS may
<br>&gt; &quot;possibly&quot; lead to remote code execution are making such a statement for a<br>&gt; couple different reasons:<br>&gt; <br>&gt; 1) They found a DoS and truly have no idea whether or not it can cause<br>&gt; remote code execution due to not having the knowledge/skills necessary to
<br>&gt; check for it and/or lack of time to make such a determination.<br>&gt; <br>&gt; 2) They have seen characteristics that would indicate that remote code<br>&gt; execution is possible but have not quite been able to nail down a working
<br>&gt; exploit &quot;should&quot; one be possible.<br>&gt; <br>&gt; I do not think the evidence quickly available to us would bring us to<br>&gt; conclude most DoS&#39;s end up resulting in remote code execution -- or even
<br>&gt; have the ability to.&nbsp;&nbsp;I would agree saying &quot;often enough&quot; would be better<br>&gt; than &quot;most.&quot;<br>&gt; <br>&gt; However, regardless of whether it results in remote code execution, I<br>&gt; don&#39;t think a DoS should necessarily be discounted as frivolous or
<br>&gt; irrelevant.&nbsp;&nbsp;It might not rank up there with critical or high<br>&gt; vulnerabilities, but it is a vulnerability nonetheless.<br>&gt; <br>&gt; Steven<br>&gt; <a href="http://securityzone.org">securityzone.org</a>
<br>&gt; <br>&gt; &gt; Ok &#39;most&#39; is probably bad wording on my part how does &#39;often enough&#39; sound<br>&gt; &gt; :).<br>&gt; &gt;<br>&gt; &gt; &quot;Buffer overflow in the png_decompress_chunk function in pngrutil.c
 in<br>&gt; &gt; libpng before 1.2.12 allows context-dependent attackers to cause a<br>&gt; &gt; denial of service and possibly execute arbitrary code&quot;<br>&gt; &gt; <a href="http://www.securityspace.com/smysecure/catid.html?id=57643">
http://www.securityspace.com/smysecure/catid.html?id=57643</a><br>&gt; &gt;<br>&gt; &gt; &quot;Buffer overflow in efingerd 1.5 and earlier, and possibly up to 1.61,<br>&gt; &gt; allows remote attackers to cause a denial of service and possibly
<br>&gt; &gt; execute arbitrary code via a finger request from an IP address with a<br>&gt; &gt; long hostname that is obtained via a reverse DNS lookup.&quot;<br>&gt; &gt; <a href="http://cve.mitre.org/board/archives/2003-03/msg00013.html">
http://cve.mitre.org/board/archives/2003-03/msg00013.html</a><br>&gt; &gt;<br>&gt; &gt; &quot;A BrightStor ARCserve Backup contains four<br>&gt; &gt; vulnerabilities that can allow a remote attacker to cause a denial<br>&gt; &gt; of service or possibly execute arbitrary code.&quot;
<br>&gt; &gt; <a href="http://packetstorm.linuxsecurity.com/0703-advisories/CAID-McAfee.txt">http://packetstorm.linuxsecurity.com/0703-advisories/CAID-McAfee.txt</a><br>&gt; &gt;<br>&gt; &gt;<br>&gt; &gt; Note the use of &#39;possibly&#39;. If it was possible then &#39;possibly&#39; wouldn&#39;t be
<br>&gt; &gt; used.<br>&gt; &gt;<br>&gt; &gt; I&#39;m not going to debate the validity of the month of activex bugs because<br>&gt; &gt; frankly I don&#39;t care, merely<br>&gt; &gt; that a DOS can turn out to be more and that at times either the researcher
<br>&gt; &gt; hasn&#39;t spent enough time on it, can&#39;t get the POC working, or lacks the<br>&gt; &gt; skill to fully understand the problem.<br>&gt; &gt;<br>&gt; &gt; There have been multiple instances on the securityfocus lists throughout
<br>&gt; &gt; the years where a DOS suddenly<br>&gt; &gt; became promoted to a remotely exploitable bug (i.e another person found it<br>&gt; &gt; was actually exploitable). I&#39;m not going<br>&gt; &gt; to find them and post them here, but a little googling can yield
<br>&gt; &gt; results.<br>&gt; &gt;<br>&gt; &gt; - Robert<br>&gt; &gt; <a href="http://www.cgisecurity.com/">http://www.cgisecurity.com/</a><br>&gt; &gt;<br>&gt; &gt;&gt; &gt;&gt;Consider that most often a bug filed as DOS can actually be
<br>&gt; &gt;&gt; exploitable, but the person who discovered it can&#39;t get the POC working<br>&gt; &gt;&gt; or is even aware it is. While command execution is the ideal goal it<br>&gt; &gt;&gt; doesn&#39;t mean other types of issues are *completely* worthless.&nbsp;&nbsp;=20
<br>&gt; &gt;&gt;<br>&gt; &gt;&gt; Most often? How do you know that?<br>&gt; &gt;&gt;<br>&gt; &gt;&gt; Larry Seltzer<br>&gt; &gt;&gt; eWEEK.com Security Center Editor<br>&gt; &gt;&gt; <a href="http://security.eweek.com/">
http://security.eweek.com/</a><br>&gt; &gt;&gt; <a href="http://blogs.eweek.com/cheap_hack/">http://blogs.eweek.com/cheap_hack/</a><br>&gt; &gt;&gt; Contributing Editor, PC Magazine<br>&gt; &gt;&gt; larryseltzer@ziffdavis.com
=20<br>&gt; &gt;&gt;<br>&gt; &gt;<br>&gt; &gt; _______________________________________________<br>&gt; &gt; Full-Disclosure - We believe in it.<br>&gt; &gt; Charter: <a href="http://lists.grok.org.uk/full-disclosure-charter.html">
http://lists.grok.org.uk/full-disclosure-charter.html</a><br>&gt; &gt; Hosted and sponsored by Secunia - <a href="http://secunia.com/">http://secunia.com/</a><br>&gt; &gt;<br>&gt; <br>&gt; <br>&gt; _______________________________________________
<br>&gt; Full-Disclosure - We believe in it.<br>&gt; Charter: <a href="http://lists.grok.org.uk/full-disclosure-charter.html">http://lists.grok.org.uk/full-disclosure-charter.html</a><br>&gt; Hosted and sponsored by Secunia - 
<a href="http://secunia.com/">http://secunia.com/</a><br>&gt; <br><br><br>-- <br><a href="http://www.goldwatches.com/watches.asp?Brand=39">http://www.goldwatches.com/watches.asp?Brand=39</a><br><a href="http://www.wazoozle.com">
http://www.wazoozle.com</a><br>