<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
  <head>
    <meta content="text/html; charset=UTF-8" http-equiv="Content-Type">
  </head>
  <body bgcolor="#ffffff" text="#000000">
    Jed,<br>
    <br>
    Thanks. However:<br>
    <br>
    [ 19%] Building C object
    CMakeFiles/petsc.dir/src/sys/ftn-custom/zstartf.c.o<br>
    [ 19%] Building C object
    CMakeFiles/petsc.dir/src/sys/ftn-custom/zsys.c.o<br>
    Building C object CMakeFiles/petsc.dir/src/sys/ftn-custom/zutils.c.o<br>
    /panfs/panfs.gfz-hpcc.cluster/home/mt/agrayver/lib/petsc-dev1/src/sys/ftn-custom/zstart.c(117):
    error: expected a type specifier<br>
      extern MPI_Op MPIU_SUM;<br>
                    ^<br>
    /panfs/panfs.gfz-hpcc.cluster/home/mt/agrayver/lib/petsc-dev1/src/sys/ftn-custom/zstart.c(117):
    error: "MPI_Op" has already been declared in the current scope<br>
      extern MPI_Op MPIU_SUM;<br>
                    ^<br>
    /panfs/panfs.gfz-hpcc.cluster/home/mt/agrayver/lib/petsc-dev1/src/sys/ftn-custom/zstart.c(128):
    error: function "MPI_Op" is not a type name<br>
      extern  MPI_Op PetscMaxSum_Op;<br>
              ^<br>
    /panfs/panfs.gfz-hpcc.cluster/home/mt/agrayver/lib/petsc-dev1/src/sys/ftn-custom/zstart.c(220):
    error: function "MPI_Op" is not a type name<br>
      extern  MPI_Op PetscADMax_Op;<br>
              ^<br>
    /panfs/panfs.gfz-hpcc.cluster/home/mt/agrayver/lib/petsc-dev1/src/sys/ftn-custom/zstart.c(221):
    error: function "MPI_Op" is not a type name<br>
      extern  MPI_Op PetscADMin_Op;<br>
              ^<br>
    /panfs/panfs.gfz-hpcc.cluster/home/mt/agrayver/lib/petsc-dev1/src/sys/ftn-custom/zstart.c(343):
    error: expression must be an lvalue or a function designator<br>
        *ierr = MPI_Op_create(PetscSum_Local,1,&MPIU_SUM);<br>
    <br>
    On 23.04.2012 03:23, Jed Brown wrote:
    <blockquote
cite="mid:CAM9tzSmP8G9MjBa1JHsuCQ32NvxdDXv8k=j3hH4fGa-7YWHj-w@mail.gmail.com"
      type="cite">Thanks for the test case. Fixed here:
      <div><br>
      </div>
      <div><a moz-do-not-send="true"
          href="http://petsc.cs.iit.edu/petsc/petsc-dev/rev/7ac7040f19d7">http://petsc.cs.iit.edu/petsc/petsc-dev/rev/7ac7040f19d7</a><br>
        <br>
        <div class="gmail_quote">On Sun, Apr 22, 2012 at 08:11,
          Alexander Grayver <span dir="ltr"><<a
              moz-do-not-send="true"
              href="mailto:agrayver@gfz-potsdam.de">agrayver@gfz-potsdam.de</a>></span>
          wrote:<br>
          <blockquote class="gmail_quote" style="margin: 0pt 0pt 0pt
            0.8ex; border-left: 1px solid rgb(204, 204, 204);
            padding-left: 1ex;">
            <div bgcolor="#ffffff" text="#000000"> Matt,<br>
              <br>
              Here are matrices (double complex, 3 MB each):<br>
              <a moz-do-not-send="true"
                href="http://dl.dropbox.com/u/60982984/X.dat"
                target="_blank">http://dl.dropbox.com/u/60982984/X.dat</a><br>
              <a moz-do-not-send="true"
                href="http://dl.dropbox.com/u/60982984/Y.dat"
                target="_blank">http://dl.dropbox.com/u/60982984/Y.dat</a><br>
              <br>
              Code:<br>
              <br>
              static char help[] = "";<br>
              int main(int argc,char **args)<br>
              {<br>
                Mat            A,B;<br>
                PetscErrorCode ierr;<br>
                PetscViewer    viewer;<br>
                PetscScalar    cone;<br>
              <br>
                PetscInitialize(&argc,&args,(char *)0,help);<br>
              <br>
                ierr =
PetscViewerBinaryOpen(PETSC_COMM_WORLD,"Y.dat",FILE_MODE_READ,&viewer);CHKERRQ(ierr);<br>
                ierr = MatCreate(PETSC_COMM_WORLD,&A);CHKERRQ(ierr);<br>
                ierr = MatSetFromOptions(A);CHKERRQ(ierr);<br>
                ierr = MatLoad(A,viewer);CHKERRQ(ierr);<br>
                ierr = PetscViewerDestroy(&viewer);CHKERRQ(ierr);<br>
              <br>
                ierr =
PetscViewerBinaryOpen(PETSC_COMM_WORLD,"X.dat",FILE_MODE_READ,&viewer);CHKERRQ(ierr);<br>
                ierr = MatCreate(PETSC_COMM_WORLD,&B);CHKERRQ(ierr);<br>
                ierr = MatSetFromOptions(B);CHKERRQ(ierr);<br>
                ierr = MatLoad(B,viewer);CHKERRQ(ierr);<br>
                ierr = PetscViewerDestroy(&viewer);CHKERRQ(ierr);<br>
                <br>
                cone = 1.0;<br>
                ierr = MatAYPX(A,cone,B,DIFFERENT_NONZERO_PATTERN);<br>
                ierr = PetscFinalize();<br>
                return 0;<br>
              }<br>
              <br>
              It works fine if you run it with:<br>
              mpirun -n 10 solveTest -mat_type aij<br>
              <br>
              and crashes if n > 10<br>
              <br>
              Thanks.
              <div>
                <div class="h5"><br>
                  <br>
                  On 22.04.2012 14:51, Matthew Knepley wrote:
                  <blockquote type="cite">On Sun, Apr 22, 2012 at 8:41
                    AM, Alexander Grayver <span dir="ltr"><<a
                        moz-do-not-send="true"
                        href="mailto:agrayver@gfz-potsdam.de"
                        target="_blank">agrayver@gfz-potsdam.de</a>></span>
                    wrote:<br>
                    <div class="gmail_quote">
                      <blockquote class="gmail_quote" style="margin: 0pt
                        0pt 0pt 0.8ex; border-left: 1px solid rgb(204,
                        204, 204); padding-left: 1ex;">
                        <div bgcolor="#ffffff" text="#000000"> Sorry, I
                          don't quite get. Is this something I can avoid
                          by doing preallocation properly or petsc issue
                          that will be fixed? </div>
                      </blockquote>
                      <div><br>
                      </div>
                      <div>This looks like a PETSc bug to me, but it
                        would be nice to get input that triggers it.</div>
                      <div><br>
                      </div>
                      <div>  Thanks,</div>
                      <div><br>
                      </div>
                      <div>     Matt</div>
                      <div> </div>
                      <blockquote class="gmail_quote" style="margin: 0pt
                        0pt 0pt 0.8ex; border-left: 1px solid rgb(204,
                        204, 204); padding-left: 1ex;">
                        <div bgcolor="#ffffff" text="#000000">
                          <div>
                            <div> On 20.04.2012 14:23, Jed Brown wrote:
                              <blockquote type="cite">
                                <div class="gmail_quote">On Fri, Apr 20,
                                  2012 at 05:18, Matthew Knepley <span
                                    dir="ltr"><<a
                                      moz-do-not-send="true"
                                      href="mailto:knepley@gmail.com"
                                      target="_blank">knepley@gmail.com</a>></span>
                                  wrote:<br>
                                  <blockquote class="gmail_quote"
                                    style="margin: 0pt 0pt 0pt 0.8ex;
                                    border-left: 1px solid rgb(204, 204,
                                    204); padding-left: 1ex;">
                                    <div>Did someone intend to fix this,
                                      and then
                                      stop? MatAXPY_BasicWithPreallocation()
                                      is the same as MatAXPY_Basic().</div>
                                  </blockquote>
                                </div>
                                <div><br>
                                </div>
                                <div>Look how it's called. I don't know
                                  what's going on here.</div>
                                <div><br>
                                </div>
                                <div>    ierr =
                                  MatAXPYGetPreallocation_SeqAIJ(yy->A,xx->A,nnz_d);CHKERRQ(ierr);</div>
                                <div>    ierr =
                                  MatAXPYGetPreallocation_SeqAIJ(yy->B,xx->B,nnz_o);CHKERRQ(ierr);</div>
                                <div>    ierr =
                                  MatMPIAIJSetPreallocation(B,0,nnz_d,0,nnz_o);CHKERRQ(ierr);</div>
                                <div>    ierr =
                                  MatAXPY_BasicWithPreallocation(B,Y,a,X,str);CHKERRQ(ierr);</div>
                                <div><br>
                                </div>
                              </blockquote>
                              <br>
                              <br>
                            </div>
                          </div>
                          <span><font color="#888888">
                              <pre cols="72">-- 
Regards,
Alexander</pre>
                            </font></span></div>
                      </blockquote>
                    </div>
                    <br>
                    <br clear="all">
                    <div><br>
                    </div>
                    -- <br>
                    What most experimenters take for granted before they
                    begin their experiments is infinitely more
                    interesting than any results to which their
                    experiments lead.<br>
                    -- Norbert Wiener<br>
                  </blockquote>
                  <br>
                  <br>
                </div>
              </div>
              <span class="HOEnZb"><font color="#888888">
                  <pre cols="72">-- 
Regards,
Alexander</pre>
                </font></span></div>
          </blockquote>
        </div>
        <br>
      </div>
    </blockquote>
    <br>
    <br>
    <pre class="moz-signature" cols="72">-- 
Regards,
Alexander</pre>
  </body>
</html>