<html>
  <head>
    <meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
  </head>
  <body>
    <p>Hi!</p>
    <p>I have been playing with oss-fuzz project over one week. I think
      many of them might be invalid, because failures are caused by
      wrong fuzzing. More precisely by incomplete initialization used
      when fuzzing. I have created fix for one [1]. I have attached
      patch, which seems prevents such failures. I am not 100% sure
      resize_packet should never increase udp message to larger packet
      than received. But because it does not have other limit available
      but plen, I used that as a top. I am confident that is correct
      limit of usable buffer in handling tcp response.<br>
    </p>
    <p>But I think <span id="summary_container"><span
          id="short_desc_nonedit_display"><a
            href="https://access.redhat.com/security/cve/CVE-2021-45955">CVE-2021-45955</a>
          might be a valid one. It seems no proper bound is checked on
          pseudo header reinsertion. Patch attached.<br>
        </span></span></p>
    <p><span id="summary_container"><span
          id="short_desc_nonedit_display">My attempts to build fuzzers
          with debuggable code were partially successful. I have pushed
          the code I use for started fuzzing at oss-fuzz branch [2]. I
          just source fuzz/env-rpm.sh, then fuzz/build.sh to create
          fuzzers.<br>
        </span></span></p>
    <p><span id="summary_container"><span
          id="short_desc_nonedit_display">It seems all functions
          crashing in extract_name are invalid, because too small buffer
          is used in fuzzer. And it correctly detects it would write
          behind allocated space. I haven't met them after [1] were
          applied.<br>
        </span></span></p>
    <p><span id="summary_container"><span
          id="short_desc_nonedit_display">Should I create better
          integration to dnsmasq upstream project? It seems to be
          interesting way of checking possible inputs to dnsmasq. Has
          anyone other been successful in fuzzing something themselves?
          Have you been able to validate details using reproducers?<br>
        </span></span></p>
    <p><span id="summary_container"><span
          id="short_desc_nonedit_display">Cheers,<br>
          Petr<br>
        </span></span></p>
    <p>1. <a class="moz-txt-link-freetext" href="https://github.com/google/oss-fuzz/pull/7293">https://github.com/google/oss-fuzz/pull/7293</a><br>
      2.
      <a class="moz-txt-link-freetext" href="https://github.com/InfrastructureServices/dnsmasq/tree/oss-fuzz/fuzz">https://github.com/InfrastructureServices/dnsmasq/tree/oss-fuzz/fuzz</a><br>
    </p>
    <div class="moz-cite-prefix">On 2/14/22 23:32, Hauke Mehrtens wrote:<br>
    </div>
    <blockquote type="cite"
      cite="mid:f3f38af6-8c3c-a2bd-051f-ad4f249076a1@hauke-m.de">Hi,
      <br>
      <br>
      Our CVE checking scripts in OpenWrt found the following recently
      opened CVEs against dnsmasq:
      <br>
      <a class="moz-txt-link-freetext" href="https://nvd.nist.gov/vuln/detail/CVE-2021-45951">https://nvd.nist.gov/vuln/detail/CVE-2021-45951</a>
      <br>
      <a class="moz-txt-link-freetext" href="https://nvd.nist.gov/vuln/detail/CVE-2021-45952">https://nvd.nist.gov/vuln/detail/CVE-2021-45952</a>
      <br>
      <a class="moz-txt-link-freetext" href="https://nvd.nist.gov/vuln/detail/CVE-2021-45953">https://nvd.nist.gov/vuln/detail/CVE-2021-45953</a>
      <br>
      <a class="moz-txt-link-freetext" href="https://nvd.nist.gov/vuln/detail/CVE-2021-45954">https://nvd.nist.gov/vuln/detail/CVE-2021-45954</a>
      <br>
      <a class="moz-txt-link-freetext" href="https://nvd.nist.gov/vuln/detail/CVE-2021-45955">https://nvd.nist.gov/vuln/detail/CVE-2021-45955</a>
      <br>
      <a class="moz-txt-link-freetext" href="https://nvd.nist.gov/vuln/detail/CVE-2021-45956">https://nvd.nist.gov/vuln/detail/CVE-2021-45956</a>
      <br>
      <a class="moz-txt-link-freetext" href="https://nvd.nist.gov/vuln/detail/CVE-2021-45957">https://nvd.nist.gov/vuln/detail/CVE-2021-45957</a>
      <br>
      <br>
      We think these CVE reports are wrong and should get rejected.
      <br>
    </blockquote>
    Not all of them. How were they validated? How do you know they are
    wrong? Have you reproduced and debugged them?<br>
    <blockquote type="cite"
      cite="mid:f3f38af6-8c3c-a2bd-051f-ad4f249076a1@hauke-m.de">
      <br>
      Hauke<br>
    </blockquote>
    <pre class="moz-signature" cols="72">-- 
Petr Menšík
Software Engineer
Red Hat, <a class="moz-txt-link-freetext" href="http://www.redhat.com/">http://www.redhat.com/</a>
email: <a class="moz-txt-link-abbreviated" href="mailto:pemensik@redhat.com">pemensik@redhat.com</a>
PGP: DFCF908DB7C87E8E529925BC4931CA5B6C9FC5CB</pre>
  </body>
</html>