<html>
  <head>
    <meta http-equiv="Content-Type" content="text/html;
      charset=windows-1252">
  </head>
  <body>
    Barry,<br>
    I am curious about your statement "does not work generically".  If I
    compile with -fdefault-integer-8,<br>
    I would assume that this produces objects/libraries that will use
    64bit integers.  As long as I have not declared<br>
    explicit kind=4 integers, what else could go wrong.<br>
    -sanjay<br>
    <br>
    PS: I am not advocating this as a great idea, but I am curious if
    there or other obscure compiler level things that could go wrong. <br>
    <br>
    <br>
    <div class="moz-cite-prefix">On 3/22/21 8:53 PM, Barry Smith wrote:<br>
    </div>
    <blockquote type="cite"
      cite="mid:F757BD03-2247-46D3-BEBF-D0AE2F469D3B@petsc.dev">
      <meta http-equiv="Content-Type" content="text/html;
        charset=windows-1252">
      <br class="">
      <div><br class="">
        <blockquote type="cite" class="">
          <div class="">On Mar 22, 2021, at 3:24 PM, Junchao Zhang <<a
              href="mailto:junchao.zhang@gmail.com" class=""
              moz-do-not-send="true">junchao.zhang@gmail.com</a>>
            wrote:</div>
          <br class="Apple-interchange-newline">
          <div class="">
            <div dir="ltr" class="">
              <div dir="ltr" class=""><br class="">
                <br class="">
              </div>
              <br class="">
              <div class="gmail_quote">
                <div dir="ltr" class="gmail_attr">On Mon, Mar 22, 2021
                  at 1:39 PM Barry Smith <<a
                    href="mailto:bsmith@petsc.dev" class=""
                    moz-do-not-send="true">bsmith@petsc.dev</a>>
                  wrote:<br class="">
                </div>
                <blockquote class="gmail_quote" style="margin:0px 0px
                  0px 0.8ex;border-left:1px solid
                  rgb(204,204,204);padding-left:1ex">
                  <div style="overflow-wrap: break-word;" class="">
                    <div class=""><br class="">
                    </div>
                       Version of PETSc and MUMPS? We fixed a bug in
                    MUMPs a couple years ago that produced error
                    messages as below. Please confirm you are using the
                    latest PETSc and MUMPS. 
                    <div class=""><br class="">
                    </div>
                    <div class="">   You can run your production version
                      with the option -malloc_debug ; this will slow it
                      down a bit but if there is memory corruption it
                      may detect it and indicate the problematic error. </div>
                    <div class=""><br class="">
                    </div>
                    <div class="">    One also has to be careful about
                      the size of the problem passed to MUMPs since
                      PETSc/MUMPs does not fully support using all 64
                      bit integers. Is it only crashing for problems
                      near 2 billion entries in the sparse matrix?</div>
                  </div>
                </blockquote>
                <div class=""> "problems near 2 billion entries"?  I
                  don't understand. Should not be an issue if building
                  petsc with 64-bit indices.</div>
              </div>
            </div>
          </div>
        </blockquote>
        <div><br class="">
        </div>
          MUMPS does not have proper support for 64 bit indices. It
        relies on add-hoc Fortran compiler command line options to
        support to converting integer to 64 bit integers and does not
        work generically. Yes, Fortran lovers have been doing this for
        30 years inside their applications but it does not really work
        in a library environment. But then a big feature of Fortran is
        "who needs libraries, we just write all the code we need"
        (except Eispack,Linpack,LAPACK :=-).</div>
      <div><br class="">
        <blockquote type="cite" class="">
          <div class="">
            <div dir="ltr" class="">
              <div class="gmail_quote">
                <div class=""><br class="">
                </div>
                <blockquote class="gmail_quote" style="margin:0px 0px
                  0px 0.8ex;border-left:1px solid
                  rgb(204,204,204);padding-left:1ex">
                  <div style="overflow-wrap: break-word;" class="">
                    <div class=""><br class="">
                    </div>
                    <div class="">     valgrind is the gold standard for
                      detecting memory corruption. </div>
                    <div class=""><br class="">
                    </div>
                    <div class="">Barry</div>
                    <div class=""><br class="">
                      <div class=""><br class="">
                        <blockquote type="cite" class="">
                          <div class="">On Mar 22, 2021, at 12:56 PM,
                            Chris Hewson <<a
                              href="mailto:chris@resfrac.com"
                              target="_blank" class=""
                              moz-do-not-send="true">chris@resfrac.com</a>>
                            wrote:</div>
                          <br class="">
                          <div class="">
                            <div dir="ltr" class="">Hi All,
                              <div class=""><br class="">
                              </div>
                              <div class="">I have been having a problem
                                with MUMPS randomly crashing in our
                                program and causing the entire program
                                to crash. I am compiling in -O2
                                optimization mode and using
                                --download-mumps etc. to compile PETSc.
                                If I rerun the program, 95%+ of the time
                                I can't reproduce the error. It seems to
                                be a similar issue to this thread:</div>
                              <div class=""><br class="">
                              </div>
                              <div class=""><a
href="https://lists.mcs.anl.gov/pipermail/petsc-users/2018-October/036372.html"
                                  target="_blank" class=""
                                  moz-do-not-send="true">https://lists.mcs.anl.gov/pipermail/petsc-users/2018-October/036372.html</a></div>
                              <div class=""><br class="">
                              </div>
                              <div class="">Similar to the resolution
                                there I am going to try and increase
                                icntl_14 and see if that resolves the
                                issue. Any other thoughts on this?</div>
                              <div class=""><br class="">
                              </div>
                              <div class="">Thanks,</div>
                              <div class="">
                                <div class="">
                                  <div dir="ltr" class="">
                                    <div dir="ltr" class="">
                                      <div class="">
                                        <div dir="ltr" class="">
                                          <div class="">
                                            <div dir="ltr" class=""><b
                                                class=""><br class="">
                                              </b></div>
                                            <div dir="ltr" class=""><b
                                                class="">Chris Hewson</b>
                                              <div class="">Senior
                                                Reservoir Simulation
                                                Engineer</div>
                                              <div class="">ResFrac</div>
                                              <div class="">+1.587.575.9792</div>
                                            </div>
                                          </div>
                                        </div>
                                      </div>
                                    </div>
                                  </div>
                                </div>
                              </div>
                            </div>
                          </div>
                        </blockquote>
                      </div>
                      <br class="">
                    </div>
                  </div>
                </blockquote>
              </div>
            </div>
          </div>
        </blockquote>
      </div>
      <br class="">
    </blockquote>
    <br>
  </body>
</html>