<html>
  <head>
    <meta content="text/html; charset=UTF-8" http-equiv="Content-Type">
  </head>
  <body bgcolor="#FFFFFF" text="#000000">
    <div class="moz-cite-prefix"><font face="Times New Roman, Times,
        serif">Hi Ludwig,</font><br>
      <br>
      <blockquote><font face="Times New Roman, Times, serif">You are
          right, the test case only checks that trimming is still
          working <span class="moz-smiley-s6"><span> :-[ </span></span>.<br>
          <br>
          The problem is that the failing thread ends with an invalid
          pointer and SIGSEV.<br>
          I imagine two possibilities:<br>
        </font>
        <ul>
          <li><font face="Times New Roman, Times, serif">The thread
              dereference a pointer that is not a valid address.<br>
              The stacktrace makes think we are in that case. The
              pointer is said to be e=0x105.<br>
              Now I do not understand how this can occur because this
              pointer is an entry returned from an internal search so it
              should be an address on the heap.<br>
              This pointer value is displayed as arg of
              slapi_entry_attr_find but can not be retrieve from the
              calling function. So I am unsure if the displayed value
              was the effective value. </font><br>
            <blockquote><tt>Thread 1 (Thread 0x7f83a2cb2700 (LWP
                30103)):</tt><tt><br>
              </tt><tt>#0  attrlist_find (a=0x105, type=0x7f8395818716
                "changetime") at ldap/servers/slapd/attrlist.c:144</tt><tt><br>
              </tt><tt>No locals.</tt><tt><br>
              </tt><tt>#1  0x00007f83a286bec2 in slapi_entry_attr_find (<b>e=0x105</b>,
                type=0x7f8395818716 "changetime", a=0x7f83a2cb1d58) at
                ldap/servers/slapd/entry.c:2362</tt><tt><br>
              </tt><tt>        r = 1547436624</tt><tt><br>
              </tt><tt>#2  0x00007f83958176b0 in trim_changelog () at
                ldap/servers/plugins/retrocl/retrocl_trim.c:287</tt><tt><br>
              </tt><tt>        sval = 0x0</tt><tt><br>
              </tt><tt>        val = 0x0</tt><tt><br>
              </tt><tt>        did_delete = 0</tt><tt><br>
              </tt><tt>        attr = 0x0</tt><tt><br>
              </tt><tt>        ldrc = &lt;optimized out&gt;</tt><tt><br>
              </tt><tt>        me = 172800</tt><tt><br>
              </tt><tt>        done = 0</tt><tt><br>
              </tt><tt>        <b>e = &lt;optimized out&gt;</b></tt></blockquote>
            <font face="Times New Roman, Times, serif">If the value
              returned by internal search was 0x105, then I think an
              error (heap corruption ?) occurred down in the internal
              search and the current stacktrace can not help us to find
              it.</font><br>
            <br>
          </li>
          <li><font face="Times New Roman, Times, serif">The thread
              dereference a pointer that is no longer allocated.<br>
              <br>
              This is the purpose of this fix as we know that the
              returned entry was accessed outside of the callback
              function, where the entry has been freed.<br>
              This fix has no test case because to trigger the crash the
              entry must be freed/allocated before the trimming thread
              access it.</font><br>
            <br>
          </li>
        </ul>
        <font face="Times New Roman, Times, serif">The fix changed the
          trimming algo so the test case I made was to verify that the
          trimming was still working.<br>
          Martin explained me that the crash occurred while there was no
          ldap client activity. I will then setup some test machines,
          run the freeipa acceptance and let them idle other the week
          end. I will run one machine without the fix and the others
          with the fix. Also I will reduce the timming period. </font><br>
        <br>
        thanks<br>
        thierry<br>
        <br>
      </blockquote>
      On 08/21/2014 04:23 PM, Ludwig Krispenz wrote:<br>
    </div>
    <blockquote cite="mid:53F600E9.40500@redhat.com" type="cite">
      <meta content="text/html; charset=UTF-8" http-equiv="Content-Type">
      The test case you provide looks like it is testing that changelog
      trimming is working, shouldn't the test case for the ticket
      produce the crash it fixes (or does it and I missed it) ?<br>
      <div class="moz-cite-prefix">On 08/21/2014 03:40 PM, thierry
        bordaz wrote:<br>
      </div>
      <blockquote cite="mid:53F5F6BD.7030507@redhat.com" type="cite">
        <meta http-equiv="content-type" content="text/html;
          charset=UTF-8">
        <font face="Times New Roman, Times, serif">Ticket <a
            moz-do-not-send="true" class="moz-txt-link-freetext"
            href="https://fedorahosted.org/389/ticket/47871">https://fedorahosted.org/389/ticket/47871</a><br>
          Fix: <a moz-do-not-send="true" class="moz-txt-link-freetext"
href="https://fedorahosted.org/389/attachment/ticket/47871/0001-Ticket-47871-389-ds-base-1.3.2.21-1.fc20-crashed-ove.patch">https://fedorahosted.org/389/attachment/ticket/47871/0001-Ticket-47871-389-ds-base-1.3.2.21-1.fc20-crashed-ove.patch</a><br>
          Test case: <a moz-do-not-send="true"
            class="moz-txt-link-freetext"
href="https://fedorahosted.org/389/attachment/ticket/47871/0002-Ticket-47871-Test-case-389-ds-base-1.3.2.21-1.fc20-c.patch">https://fedorahosted.org/389/attachment/ticket/47871/0002-Ticket-47871-Test-case-389-ds-base-1.3.2.21-1.fc20-c.patch</a><br>
          <br>
          Thanks<br>
          thierry<br>
        </font> <br>
        <fieldset class="mimeAttachmentHeader"></fieldset>
        <br>
        <pre wrap="">--
389-devel mailing list
<a moz-do-not-send="true" class="moz-txt-link-abbreviated" href="mailto:389-devel@lists.fedoraproject.org">389-devel@lists.fedoraproject.org</a>
<a moz-do-not-send="true" class="moz-txt-link-freetext" href="https://admin.fedoraproject.org/mailman/listinfo/389-devel">https://admin.fedoraproject.org/mailman/listinfo/389-devel</a></pre>
      </blockquote>
      <br>
      <br>
      <fieldset class="mimeAttachmentHeader"></fieldset>
      <br>
      <pre wrap="">--
389-devel mailing list
<a class="moz-txt-link-abbreviated" href="mailto:389-devel@lists.fedoraproject.org">389-devel@lists.fedoraproject.org</a>
<a class="moz-txt-link-freetext" href="https://admin.fedoraproject.org/mailman/listinfo/389-devel">https://admin.fedoraproject.org/mailman/listinfo/389-devel</a></pre>
    </blockquote>
    <br>
  </body>
</html>