<html>
  <head>
    <meta content="text/html; charset=windows-1252"
      http-equiv="Content-Type">
  </head>
  <body bgcolor="#FFFFFF" text="#000000">
    Hi, yes you'll need the 10 bit labels if you go above 256 basis
    functions. I think I saw this error before but I don't remember what
    it was about. Does it happen with the precompiled binaries, as well?<br>
    <br>
    I can also cross-check it with the current development version.<br>
    <br>
    -Felix<br>
    <br>
    <div class="moz-cite-prefix">On 07/20/2016 09:38 PM, Christopher
      Malbon wrote:<br>
    </div>
    <blockquote
cite="mid:CAGaYXODPpa4jJnR7mXHLcfJzQ+pyX3J2NR=0OW6eO5KTaNeWyg@mail.gmail.com"
      type="cite">
      <div dir="ltr">
        <div>From $COLUMBUS/docs/fulldocs/sifs.txt:<br>
          info(6) contains information on the format of the packed
          integral labels read from the header in the integral file. <br>
          0 = 8 bits/label. 1 = 10 bits/label<br>
        </div>
        <div>ifmt is the format of the packed integral labels in the
          buffer.<br>
        </div>
        <div><br>
        </div>
        <div>Systems with basis functions >256 require >8 bits to
          encode a basis function label.<br>
        </div>
        <div>It's essential a consistency check to ensure you are
          assigning integrals the appropriate basis function labels.
          This is at least my understanding.<br>
          <br>
        </div>
        Best
        <div>
          <div class="gmail_signature" data-smartmail="gmail_signature">
            <div dir="ltr">
              <div>
                <div><br>
                  Christopher L Malbon<br>
                  Yarkony Group<br>
                </div>
                Department of Chemistry<br>
              </div>
              The Johns Hopkins University<br>
              <a moz-do-not-send="true" href="mailto:cmalbon1@jhu.edu"
                target="_blank">cmalbon1@jhu.edu</a><br>
            </div>
          </div>
        </div>
        <br>
        <div class="gmail_quote">---------- Forwarded message ----------<br>
          From: <b class="gmail_sendername">Christopher Malbon</b> <span
            dir="ltr"><<a moz-do-not-send="true"
              href="mailto:clmalbon@gmail.com">clmalbon@gmail.com</a>></span><br>
          Date: Wed, Jul 20, 2016 at 3:28 PM<br>
          Subject: Re: [Columbus] mcscf.x MO integral read error<br>
          To: Scott Brozell <<a moz-do-not-send="true"
            href="mailto:srb@osc.edu">srb@osc.edu</a>>, Christopher
          Malbon <<a moz-do-not-send="true"
            href="mailto:clmalbon@gmail.com">clmalbon@gmail.com</a>>,
          <a moz-do-not-send="true" href="mailto:columbus@osc.edu">columbus@osc.edu</a><br>
          <br>
          <br>
          <div dir="ltr">
            <div>From $COLUMBUS/docs/fulldocs/sifs.txt:<br>
              info(6) contains information on the format of the packed
              integral labels read from the header in the integral file.
              <br>
              0 = 8 bits/label. 1 = 10 bits/label<br>
            </div>
            <div>ifmt is the format of the packed integral labels in the
              buffer.<br>
            </div>
            <div><br>
            </div>
            <div>Systems with basis functions >256 require >8 bits
              to encode a basis function label.<br>
            </div>
            <div>It's essential a consistency check to ensure you are
              assigning integrals the appropriate basis function labels.
              This is at least my understanding.<br>
              <br>
            </div>
            <div>Best<br>
            </div>
          </div>
          <div class="gmail_extra"><span class=""><br clear="all">
              <div>
                <div data-smartmail="gmail_signature">
                  <div dir="ltr">
                    <div>
                      <div><br>
                        Christopher L Malbon<br>
                        Yarkony Group<br>
                      </div>
                      Department of Chemistry<br>
                    </div>
                    The Johns Hopkins University<br>
                    <a moz-do-not-send="true"
                      href="mailto:cmalbon1@jhu.edu" target="_blank">cmalbon1@jhu.edu</a><br>
                  </div>
                </div>
              </div>
              <br>
            </span>
            <div>
              <div class="h5">
                <div class="gmail_quote">On Wed, Jul 20, 2016 at 2:24
                  PM, Scott Brozell <span dir="ltr"><<a
                      moz-do-not-send="true" href="mailto:srb@osc.edu"
                      target="_blank">srb@osc.edu</a>></span> wrote:<br>
                  <blockquote class="gmail_quote" style="margin:0px 0px
                    0px 0.8ex;border-left:1px solid
                    rgb(204,204,204);padding-left:1ex">Hi,<br>
                    <br>
                    The error message is new code in file colib8
                    subroutine sifd1 of<br>
                    columbus version 7 that is not in older columbus
                    versions:<br>
                    <br>
                          call ulab16( buffer(1), unpack, 4 )<br>
                          num    = unpack(1)<br>
                          lab1   = unpack(2)<br>
                          last   = mod(unpack(4),         4)<br>
                          ifmt   = mod(unpack(4)/2**2,    8)<br>
                          itypeb = mod(unpack(3),      1024)<br>
                          itypea = mod(unpack(3)/2**10,   8)<br>
                          ibvtyp = mod(unpack(3)/2**13,   8)<br>
                    c<br>
                    c     ifmt is now double checked<br>
                    c<br>
                           if (<a moz-do-not-send="true"
                      href="http://ifmt.ne.info" rel="noreferrer"
                      target="_blank">ifmt.ne.info</a>(6)) then<br>
                            call bummer(' sifd1: <a
                      moz-do-not-send="true" href="http://ifmt.ne.info"
                      rel="noreferrer" target="_blank">ifmt.ne.info</a>(6)',0,2)<br>
                           endif<br>
                    <br>
                    The comment does not indicate why this check is
                    useful.  Perhaps the<br>
                    author of the new code can describe its purpose.<br>
                    <span><font color="#888888"><br>
                        scott<br>
                      </font></span>
                    <div>
                      <div><br>
                        On Tue, Jul 19, 2016 at 05:16:47PM -0400,
                        Christopher Malbon wrote:<br>
                        > mcscf.x routine fails with the following
                        error message:<br>
                        > end of hernew<br>
                        >  IGNORING BNDCHK<br>
                        > *** End of DALTON calculation ***<br>
                        > end of mcdrt<br>
                        > end of mcuft<br>
                        >  bummer (fatal): sifd1: <a
                          moz-do-not-send="true"
                          href="http://ifmt.ne.info" rel="noreferrer"
                          target="_blank">ifmt.ne.info</a>(6)0<br>
                        > bummer<br>
                        ><br>
                        > This bummer error is called during the read
                        of the integral records after<br>
                        > ulab16 unpacks the buffer.<br>
                        ><br>
                        > I am using the Col7.0_2015_01_15 release,
                        building from source.<br>
                        > I am using the cray intel fortran wrappers
                        on the NERSC cori machine.<br>
                        > I had to change the cpp flags omitting the
                        -C option. (In gcc 4.8.x, cpp -C<br>
                        > includes the header file stdc-predef.h
                        which prints the GNU license at the<br>
                        > top of each preprocessed file commented
                        C-style, which fortran compilers do<br>
                        > not handle.) Other than this change, no
                        other changes were made - with the<br>
                        > exception of 'ftn' and 'cc' for 'ifort' and
                        'icc', respectively.<br>
                        ><br>
                        > I am installing using
                        machine.cfg/linux64.ifc.byterecl.<br>
                      </div>
                    </div>
                  </blockquote>
                </div>
                <br>
              </div>
            </div>
          </div>
        </div>
        <br>
      </div>
      <br>
      <fieldset class="mimeAttachmentHeader"></fieldset>
      <br>
      <pre wrap="">_______________________________________________
Columbus mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Columbus@lists.osc.edu">Columbus@lists.osc.edu</a>
<a class="moz-txt-link-freetext" href="https://lists.osu.edu/mailman/listinfo/columbus">https://lists.osu.edu/mailman/listinfo/columbus</a></pre>
    </blockquote>
    <br>
    <pre class="moz-signature" cols="72">-- 
Felix Plasser
Institute for Theoretical Chemistry
University of Vienna
Währingerstr. 17/404
1090 Wien
<a class="moz-txt-link-abbreviated" href="mailto:felix.plasser@univie.ac.at">felix.plasser@univie.ac.at</a>
<a class="moz-txt-link-freetext" href="http://homepage.univie.ac.at/felix.plasser/">http://homepage.univie.ac.at/felix.plasser/</a>
Tel.: +43-1-4277-52757</pre>
  </body>
</html>