<html>
  <head>
    <meta content="text/html; charset=ISO-8859-1"
      http-equiv="Content-Type">
  </head>
  <body bgcolor="#FFFFFF" text="#000000">
    No, I use global.<br>
    <br>
    <tt>&nbsp; for(Int i=0; i&lt;n; i++) {<br>
      &nbsp;&nbsp;&nbsp; Int gi = mesh.nodes[i].global_number;<br>
      &nbsp;&nbsp;&nbsp; VecGetValues(x, 1, &amp;gi, &amp;unk[i]);<br>
      &nbsp; }<br>
    </tt><br>
    "n" is defined as the number of cells inside, i.e. without buffers.&nbsp;
    "unk" is my external array.&nbsp; If I try to access buffer values, I
    use:<br>
    <br>
    <tt>&nbsp; for(Int i=0; i&lt;N; i++) {<br>
      &nbsp;&nbsp;&nbsp; Int gi = mesh.nodes[i].global_number;<br>
      &nbsp;&nbsp;&nbsp; VecGetValues(x, 1, &amp;gi, &amp;unk[i]);<br>
      &nbsp; }<br>
    </tt><br>
    But then I end up with tons of warnings, presumably because I am
    going beyond "n".&nbsp; Vector x was created with VecCreateGhost.<br>
    <br>
    On 2/23/2012 10:00 PM, Mohammad Mirzadeh wrote:
    <blockquote
cite="mid:CAKsPKwm+kjohxzFS2PE2N1XOH1ORWyXu1o0vS+Gq9bcfpE8rZQ@mail.gmail.com"
      type="cite">
      <div dir="ltr">Are you using local numbering when accessing the
        local part of ghost nodes?<br>
        <br>
        <div class="gmail_quote">On Thu, Feb 23, 2012 at 12:43 PM, Bojan
          Niceno <span dir="ltr">&lt;<a moz-do-not-send="true"
              href="mailto:bojan.niceno@psi.ch">bojan.niceno@psi.ch</a>&gt;</span>
          wrote:<br>
          <blockquote class="gmail_quote" style="margin:0 0 0
            .8ex;border-left:1px #ccc solid;padding-left:1ex">
            <div bgcolor="#FFFFFF" text="#000000"> Dear Matt,<br>
              <br>
              <br>
              I have a new insight, although is not the full
              resolution.&nbsp; If I change my code in PETScSolver.cpp from:<br>
              <br>
              <tt><br>
                &nbsp; </tt><tt>/*-------------------------------------------------+<br>
                &nbsp; |&nbsp; Make necessary PETSc intializations for vetors&nbsp; |<br>
                &nbsp; +-------------------------------------------------*/<br>
                &nbsp; Int&nbsp;&nbsp; nghost = N - n;<br>
                &nbsp; Int * ghosts = new Int(nghost);<br>
                &nbsp; for(Int n=0; n&lt;M.mesh.nodes.size(); n++) {<br>
                &nbsp;&nbsp;&nbsp; assert( M.mesh.nodes[n].global_number &gt;= 0);<br>
                &nbsp;&nbsp;&nbsp; assert( M.mesh.nodes[n].global_number &lt; 14065);<br>
                &nbsp; }<br>
                &nbsp; for(Int i=n; i&lt;N; i++) {<br>
                &nbsp;&nbsp;&nbsp; assert( M.mesh.nodes[i].global_number &gt;= 0);<br>
                &nbsp;&nbsp;&nbsp; assert( M.mesh.nodes[i].global_number &lt; 14065);<br>
                &nbsp;&nbsp;&nbsp; assert( ! (M.mesh.nodes[i].global_number &gt;=
                n_start &amp;&amp;<br>
                &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; M.mesh.nodes[i].global_number &lt;&nbsp;
                n_end) );<br>
                &nbsp;&nbsp;&nbsp;&nbsp;&nbsp; ghosts[i] = M.mesh.nodes[i].global_number;<br>
                &nbsp; }<br>
                <br>
                &nbsp; VecCreateGhost(PETSC_COMM_WORLD, n, PETSC_DECIDE,
                nghost, &amp;ghosts[0], &amp;x);<br>
              </tt><br>
              to:<br>
              <br>
              <tt>&nbsp; /*-------------------------------------------------+<br>
                &nbsp; |&nbsp; Make necessary PETSc intializations for vetors&nbsp; |<br>
                &nbsp; +-------------------------------------------------*/<br>
                &nbsp; Int&nbsp;&nbsp;&nbsp;&nbsp; nghost = N - n;<br>
                &nbsp; Indices
                ghosts;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; //
                &lt;---= NEW!<br>
                &nbsp; for(Int n=0; n&lt;M.mesh.nodes.size(); n++) {<br>
                &nbsp;&nbsp;&nbsp; assert( M.mesh.nodes[n].global_number &gt;= 0);<br>
                &nbsp;&nbsp;&nbsp; assert( M.mesh.nodes[n].global_number &lt; 14065);<br>
                &nbsp; }<br>
                &nbsp; for(Int i=n; i&lt;N; i++) {<br>
                &nbsp;&nbsp;&nbsp; assert( M.mesh.nodes[i].global_number &gt;= 0);<br>
                &nbsp;&nbsp;&nbsp; assert( M.mesh.nodes[i].global_number &lt; 14065);<br>
                &nbsp;&nbsp;&nbsp; assert( ! (M.mesh.nodes[i].global_number &gt;=
                n_start &amp;&amp;<br>
                &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; M.mesh.nodes[i].global_number &lt;&nbsp;
                n_end) );<br>
                &nbsp;&nbsp;&nbsp;&nbsp;&nbsp; ghosts.push_back( M.mesh.nodes[i].global_number );</tt><tt>&nbsp;
                // &lt;---= NEW!</tt><tt>&nbsp;&nbsp;&nbsp; <br>
                &nbsp; }<br>
                &nbsp; assert( ghosts.size() == nghost
                );&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; </tt><tt>// &lt;---= NEW!</tt><br>
              <tt><br>
                &nbsp; VecCreateGhost(PETSC_COMM_WORLD, n, PETSC_DECIDE,
                nghost, &amp;ghosts[0], &amp;x);</tt><br>
              <br>
              I pass the VecCreateGhost phase.&nbsp; "Indices" is an STL
              container of integers.&nbsp; It seems it works better than
              classical C array for this case.<br>
              <br>
              <br>
              However, I still do not see the ghost values, i.e. I get
              the following error:<br>
              <br>
              <tt>[1]PETSC ERROR: --------------------- Error Message
                ------------------------------------<br>
                [1]PETSC ERROR: [0]PETSC ERROR: ---------------------
                Error Message ------------------------------------<br>
                [0]PETSC ERROR: Argument out of range!<br>
                [0]PETSC ERROR: Can only get local values, trying 3529!<br>
                [0]PETSC ERROR:
                ------------------------------------------------------------------------<br>
                [0]PETSC ERROR: Argument out of range!<br>
                [1]PETSC ERROR: Can only get local values, trying 22!<br>
                [1]PETSC ERROR:
                ------------------------------------------------------------------------<br>
                [1]PETSC ERROR: Petsc Release Version 3.2.0, Patch 6,
                Wed Jan 11 09:28:45 CST 2012<br>
                [1]PETSC ERROR: See docs/changes/index.html for recent
                updates.<br>
                [1]PETSC ERROR: See docs/faq.html for hints about
                trouble shooting.<br>
                [1]PETSC ERROR: See docs/index.html for manual pages.<br>
                [1]PETSC ERROR: [2]PETSC ERROR: ---------------------
                Error Message ------------------------------------<br>
                [2]PETSC ERROR: Argument out of range!<br>
                [2]PETSC ERROR: Can only get local values, trying 86!<br>
                [2]PETSC ERROR:
                ------------------------------------------------------------------------<br>
                [2]PETSC ERROR: Petsc Release Version 3.2.0, Patch 6,
                Wed Jan 11 09:28:45 CST 2012</tt><br>
              <br>
              when I am trying to access values in ghost cells.&nbsp; What do
              I have to use to see them ghosts?&nbsp; I reckon
              VecGhostGetLocalForm should be used, right?<br>
              <br>
              <br>
              &nbsp;&nbsp;&nbsp; Kind regards,<br>
              <br>
              <br>
              &nbsp;&nbsp;&nbsp; Bojan
              <div>
                <div class="h5"><br>
                  <br>
                  <br>
                  On 2/23/2012 8:36 PM, Matthew Knepley wrote:
                  <blockquote type="cite">On Thu, Feb 23, 2012 at 1:33
                    PM, Bojan Niceno <span dir="ltr">&lt;<a
                        moz-do-not-send="true"
                        href="mailto:bojan.niceno@psi.ch"
                        target="_blank">bojan.niceno@psi.ch</a>&gt;</span>
                    wrote:<br>
                    <div class="gmail_quote">
                      <blockquote class="gmail_quote" style="margin:0 0
                        0 .8ex;border-left:1px #ccc
                        solid;padding-left:1ex">
                        <div bgcolor="#FFFFFF" text="#000000"> Dear
                          Matt,<br>
                          <br>
                          <br>
                          I sent the code as an attached tarball.&nbsp; I
                          sent it with case I run, so is 2 MB big.&nbsp; It
                          is now in the cue for moderator's approval. <br>
                        </div>
                      </blockquote>
                      <div><br>
                      </div>
                      <div>No, you HAVE to send it to <a
                          moz-do-not-send="true"
                          href="mailto:petsc-maint@mcs.anl.gov"
                          target="_blank">petsc-maint@mcs.anl.gov</a>,
                        as I said last time, for exactly this reason.</div>
                      <div><br>
                      </div>
                      <div>&nbsp; Matt</div>
                      <div>&nbsp;</div>
                      <blockquote class="gmail_quote" style="margin:0 0
                        0 .8ex;border-left:1px #ccc
                        solid;padding-left:1ex">
                        <div bgcolor="#FFFFFF" text="#000000"> Thanks.<br>
                          <br>
                          &nbsp;&nbsp;&nbsp; <br>
                          &nbsp;&nbsp;&nbsp; Kind regards,<br>
                          <br>
                          <br>
                          &nbsp;&nbsp;&nbsp; Bojan<br>
                          <br>
                          <br>
                          On 2/23/2012 8:04 PM, Matthew Knepley wrote:
                          <blockquote type="cite">On Thu, Feb 23, 2012
                            at 12:51 PM, Bojan Niceno <span dir="ltr">&lt;<a
                                moz-do-not-send="true"
                                href="mailto:bojan.niceno@psi.ch"
                                target="_blank">bojan.niceno@psi.ch</a>&gt;</span>
                            wrote:<br>
                            <div class="gmail_quote">
                              <blockquote class="gmail_quote"
                                style="margin:0 0 0 .8ex;border-left:1px
                                #ccc solid;padding-left:1ex">
                                <div bgcolor="#FFFFFF" text="#000000">
                                  Dear Matt, <br>
                                  <br>
                                  <br>
                                  are you sure?&nbsp; It is almost 4000 lines
                                  long!&nbsp; Shall I send only the function
                                  which bother me?<br>
                                  <br>
                                  If the entire code is what you need,
                                  shall I make a tarball and attach it?<br>
                                </div>
                              </blockquote>
                              <div><br>
                              </div>
                              <div>Send something the builds and runs.
                                Don't care how long it is.</div>
                              <div><br>
                              </div>
                              <div>&nbsp; &nbsp;Matt</div>
                              <div>&nbsp;</div>
                              <blockquote class="gmail_quote"
                                style="margin:0 0 0 .8ex;border-left:1px
                                #ccc solid;padding-left:1ex">
                                <div bgcolor="#FFFFFF" text="#000000">
                                  &nbsp;&nbsp;&nbsp; Kind regards,<br>
                                  <br>
                                  <br>
                                  &nbsp;&nbsp;&nbsp; Bojan<br>
                                  <br>
                                  On 2/23/2012 7:44 PM, Matthew Knepley
                                  wrote:
                                  <blockquote type="cite">On Thu, Feb
                                    23, 2012 at 12:28 PM, Bojan Niceno <span
                                      dir="ltr">&lt;<a
                                        moz-do-not-send="true"
                                        href="mailto:bojan.niceno@psi.ch"
                                        target="_blank">bojan.niceno@psi.ch</a>&gt;</span>
                                    wrote:<br>
                                    <div class="gmail_quote">
                                      <blockquote class="gmail_quote"
                                        style="margin:0 0 0
                                        .8ex;border-left:1px #ccc
                                        solid;padding-left:1ex">
                                        <div bgcolor="#FFFFFF"
                                          text="#000000"> On 2/23/2012
                                          7:24 PM, Matthew Knepley
                                          wrote:
                                          <blockquote type="cite">On
                                            Thu, Feb 23, 2012 at 12:05
                                            PM, Bojan Niceno <span
                                              dir="ltr">&lt;<a
                                                moz-do-not-send="true"
                                                href="mailto:bojan.niceno@psi.ch"
                                                target="_blank">bojan.niceno@psi.ch</a>&gt;</span>
                                            wrote:<br>
                                            <div class="gmail_quote">
                                              <blockquote
                                                class="gmail_quote"
                                                style="margin:0 0 0
                                                .8ex;border-left:1px
                                                #ccc
                                                solid;padding-left:1ex">
                                                <div bgcolor="#FFFFFF"
                                                  text="#000000"> Dear
                                                  Matthew,<br>
                                                  <br>
                                                  <br>
                                                  thank you for your
                                                  response.&nbsp; When I use
                                                  VecCreateGhost, I get
                                                  the following:<br>
                                                </div>
                                              </blockquote>
                                              <div><br>
                                              </div>
                                              <div>It appears that you
                                                passed a bad
                                                communicator. Did you
                                                not initialize a 'comm'
                                                variable?</div>
                                            </div>
                                          </blockquote>
                                          <br>
                                          I pass PETSC_COMM_WORLD to
                                          VecCreateGhost.&nbsp; <br>
                                          <br>
                                          I don't know what you mean by
                                          'comm' variable :-(&nbsp; I called
                                          all the routines to initialize
                                          PETSc.</div>
                                      </blockquote>
                                      <div><br>
                                      </div>
                                      <div>Send your code to <a
                                          moz-do-not-send="true"
                                          href="mailto:petsc-maint@mcs.anl.gov"
                                          target="_blank">petsc-maint@mcs.anl.gov</a>.</div>
                                      <div><br>
                                      </div>
                                      <div>&nbsp; &nbsp;Matt</div>
                                      <div>&nbsp;</div>
                                      <blockquote class="gmail_quote"
                                        style="margin:0 0 0
                                        .8ex;border-left:1px #ccc
                                        solid;padding-left:1ex">
                                        <div bgcolor="#FFFFFF"
                                          text="#000000">&nbsp; &nbsp;&nbsp;<br>
                                          &nbsp;&nbsp;&nbsp; Cheers,<br>
                                          <br>
                                          &nbsp;&nbsp;&nbsp; <br>
                                          &nbsp;&nbsp;&nbsp; Bojan<br>
                                          <br>
                                          <blockquote type="cite">
                                            <div class="gmail_quote">
                                              <div><br>
                                              </div>
                                              <div>&nbsp; &nbsp;Matt</div>
                                              <div>&nbsp;</div>
                                              <blockquote
                                                class="gmail_quote"
                                                style="margin:0 0 0
                                                .8ex;border-left:1px
                                                #ccc
                                                solid;padding-left:1ex">
                                                <div bgcolor="#FFFFFF"
                                                  text="#000000">
                                                  [0]PETSC ERROR:
------------------------------------------------------------------------<br>
                                                  [0]PETSC ERROR: Caught
                                                  signal number 11 SEGV:
                                                  Segmentation
                                                  Violation, probably
                                                  memory access out of
                                                  range<br>
                                                  [0]PETSC ERROR: Try
                                                  option
                                                  -start_in_debugger or
-on_error_attach_debugger<br>
                                                  [0]PETSC ERROR: or see
                                                  <a
                                                    moz-do-not-send="true"
href="http://www.mcs.anl.gov/petsc/petsc-as/documentation/faq.html#valgrind"
                                                    target="_blank">http://www.mcs.anl.gov/petsc/petsc-as/documentation/faq.html#valgrind</a>[0]PETSC





                                                  ERROR: or try <a
                                                    moz-do-not-send="true"
href="http://valgrind.org" target="_blank">http://valgrind.org</a> on
                                                  GNU/linux and Apple
                                                  Mac OS X to find
                                                  memory corruption
                                                  errors<br>
                                                  [0]PETSC ERROR: likely
                                                  location of problem
                                                  given in stack below<br>
                                                  [0]PETSC ERROR:
                                                  ---------------------&nbsp;
                                                  Stack Frames
                                                  ------------------------------------<br>
                                                  [0]PETSC ERROR: Note:
                                                  The EXACT line numbers
                                                  in the stack are not
                                                  available,<br>
                                                  [0]PETSC ERROR:&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;
                                                  INSTEAD the line
                                                  number of the start of
                                                  the function<br>
                                                  [0]PETSC ERROR:&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;
                                                  is given.<br>
                                                  [0]PETSC ERROR: [0]
                                                  PetscCommDuplicate
                                                  line 140
                                                  src/sys/objects/tagm.c<br>
                                                  [0]PETSC ERROR: [0]
                                                  PetscHeaderCreate_Private
                                                  line 30
                                                  src/sys/objects/inherit.c<br>
                                                  [0]PETSC ERROR: [0]
                                                  VecCreate line 32
                                                  src/vec/vec/interface/veccreate.c<br>
                                                  [0]PETSC ERROR: [0]
                                                  VecCreateGhostWithArray
                                                  line 567
                                                  src/vec/vec/impls/mpi/pbvec.c<br>
                                                  [0]PETSC ERROR: [0]
                                                  VecCreateGhost line
                                                  647
                                                  src/vec/vec/impls/mpi/pbvec.c<br>
                                                  [0]PETSC ERROR:
                                                  ---------------------
                                                  Error Message
                                                  ------------------------------------<br>
                                                  [0]PETSC ERROR: Signal
                                                  received!<br>
                                                  [0]PETSC ERROR:
                                                  ------------------------------------------------------------------------<br>
                                                  [0]PETSC ERROR: Petsc
                                                  Release Version 3.2.0,
                                                  Patch 6, Wed Jan 11
                                                  09:28:45 CST 2012<br>
                                                  [0]PETSC ERROR: See
                                                  docs/changes/index.html
                                                  for recent updates.<br>
                                                  [0]PETSC ERROR: See
                                                  docs/faq.html for
                                                  hints about trouble
                                                  shooting.<br>
                                                  [0]PETSC ERROR: See
                                                  docs/index.html for
                                                  manual pages.<br>
                                                  [0]PETSC ERROR:
                                                  ------------------------------------------------------------------------<br>
                                                  [0]PETSC ERROR:
                                                  ./PSI-Flow on a
                                                  arch-linu named
                                                  lccfd06 by niceno Thu
                                                  Feb 23 19:02:45 2012<br>
                                                  [0]PETSC ERROR:
                                                  Libraries linked from
/homecfd/niceno/PETSc-3.2-p6/arch-linux2-c-debug/lib<br>
                                                  [0]PETSC ERROR:
                                                  Configure run at Fri
                                                  Feb 10 10:24:13 2012<br>
                                                  [0]PETSC ERROR:
                                                  Configure options<br>
                                                  [0]PETSC ERROR:
                                                  ------------------------------------------------------------------------<br>
                                                  [0]PETSC ERROR: User
                                                  provided function()
                                                  line 0 in unknown
                                                  directory unknown file<br>
                                                  <br>
                                                  I don't understand
                                                  what could be causing
                                                  it.&nbsp; I took very good
                                                  care to match the
                                                  global numbers of
                                                  ghost cells when
                                                  calling VecCreateGhost<br>
                                                  <br>
                                                  <br>
                                                  &nbsp;&nbsp;&nbsp; Kind regards,<br>
                                                  <br>
                                                  <br>
                                                  &nbsp;&nbsp;&nbsp; Bojan<br>
                                                  <br>
                                                  <br>
                                                  On 2/23/2012 5:53 PM,
                                                  Matthew Knepley wrote:
                                                  <blockquote
                                                    type="cite">On Thu,
                                                    Feb 23, 2012 at
                                                    10:46 AM, Bojan
                                                    Niceno <span
                                                      dir="ltr">&lt;<a
                                                        moz-do-not-send="true"
href="mailto:bojan.niceno@psi.ch" target="_blank">bojan.niceno@psi.ch</a>&gt;</span>
                                                    wrote:<br>
                                                    <div
                                                      class="gmail_quote">
                                                      <blockquote
                                                        class="gmail_quote"
                                                        style="margin:0
                                                        0 0
                                                        .8ex;border-left:1px
                                                        #ccc
                                                        solid;padding-left:1ex">
                                                        Hi all,<br>
                                                        <br>
                                                        I've never used
                                                        a mailing list
                                                        before, so I
                                                        hope this
                                                        message will
                                                        reach PETSc
                                                        users and
                                                        experts and
                                                        someone might be
                                                        willing to help
                                                        me. &nbsp;I am also
                                                        novice in PETSc.<br>
                                                        <br>
                                                        I have developed
                                                        an unstructured
                                                        finite volume
                                                        solver on top of
                                                        PETSc libraries.
                                                        &nbsp;In sequential,
                                                        it works like a
                                                        charm. &nbsp;For the
                                                        parallel
                                                        version, I do
                                                        domain
                                                        decomposition
                                                        externally with
                                                        Metis, and work
                                                        out local and
                                                        global
                                                        numberings, as
                                                        well as
                                                        communication
                                                        patterns between
                                                        processor. &nbsp;(The
                                                        latter don't
                                                        seem to be
                                                        needed for
                                                        PETSc, though.)
                                                        &nbsp;When I run my
                                                        program in
                                                        parallel, it
                                                        also works, but
                                                        I miss values in
                                                        vectors' ghost
                                                        points.<br>
                                                        <br>
                                                        I create vectors
                                                        with command:
                                                        VecCreate(PETSC_COMM_WORLD,
                                                        &amp;x);<br>
                                                        <br>
                                                        Is it possible
                                                        to get the ghost
                                                        values if a
                                                        vector is
                                                        created like
                                                        this?<br>
                                                      </blockquote>
                                                      <div><br>
                                                      </div>
                                                      <div>I do not
                                                        understand this
                                                        question. By
                                                        definition,
                                                        "ghost values"
                                                        are those not
                                                        stored in the
                                                        global vector.</div>
                                                      <div>&nbsp;</div>
                                                      <blockquote
                                                        class="gmail_quote"
                                                        style="margin:0
                                                        0 0
                                                        .8ex;border-left:1px
                                                        #ccc
                                                        solid;padding-left:1ex">
                                                        I have tried to
                                                        use
                                                        VecCreateGhost,
                                                        but for some
                                                        reason which is
                                                        beyond my
                                                        comprehension,
                                                        PETSc goes
                                                        berserk when it
                                                        reaches the
                                                        command:
                                                        VecCreateGhost(PETSC_COMM_WORLD,
                                                        n, PETSC_DECIDE,
                                                        nghost, ifrom,
                                                        &amp;x)<br>
                                                      </blockquote>
                                                      <div><br>
                                                      </div>
                                                      <div>I think you
                                                        can understand
                                                        that "berserk"
                                                        tells me
                                                        absolutely
                                                        nothing. Error
                                                        message? Stack
                                                        trace? Did you
                                                        try to run an</div>
                                                      <div>example which
                                                        uses VecGhost?</div>
                                                      <div><br>
                                                      </div>
                                                      <div>&nbsp; Thanks,</div>
                                                      <div><br>
                                                      </div>
                                                      <div>&nbsp; &nbsp; &nbsp;Matt</div>
                                                      <div>&nbsp;</div>
                                                      <blockquote
                                                        class="gmail_quote"
                                                        style="margin:0
                                                        0 0
                                                        .8ex;border-left:1px
                                                        #ccc
                                                        solid;padding-left:1ex">
                                                        Can anyone help
                                                        me? &nbsp;Either how
                                                        to reach ghost
                                                        values for
                                                        vector created
                                                        by VecCreate, or
                                                        how to use
                                                        VecCreateGhost
                                                        properly?<br>
                                                        <br>
                                                        <br>
                                                        &nbsp; Kind regards,<br>
                                                        <br>
                                                        &nbsp; Bojan<br>
                                                      </blockquote>
                                                    </div>
                                                    <br>
                                                    <br clear="all">
                                                    <span><font
                                                        color="#888888">
                                                        <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>
                                                      </font></span></blockquote>
                                                  <span><font
                                                      color="#888888"> <br>
                                                      <br>
                                                      <div>-- <br>
                                                        <img
                                                          src="cid:part1.01070109.07010904@psi.ch"
                                                          border="0"></div>
                                                    </font></span></div>
                                              </blockquote>
                                            </div>
                                            <br>
                                            <br clear="all">
                                            <span><font color="#888888">
                                                <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>
                                              </font></span></blockquote>
                                          <span><font color="#888888"> <br>
                                              <br>
                                              <div>-- <br>
                                                <img
                                                  src="cid:part2.02030305.00030608@psi.ch"
                                                  border="0"></div>
                                            </font></span></div>
                                      </blockquote>
                                    </div>
                                    <br>
                                    <br clear="all">
                                    <span><font color="#888888">
                                        <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>
                                      </font></span></blockquote>
                                  <span><font color="#888888"> <br>
                                      <br>
                                      <div>-- <br>
                                        <img
                                          src="cid:part3.01010507.00050706@psi.ch"
                                          border="0"></div>
                                    </font></span></div>
                              </blockquote>
                            </div>
                            <br>
                            <br clear="all">
                            <span><font color="#888888">
                                <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>
                              </font></span></blockquote>
                          <span><font color="#888888"> <br>
                              <br>
                              <div>-- <br>
                                <img
                                  src="cid:part4.00020303.02000002@psi.ch"
                                  border="0"></div>
                            </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">
                  <div>-- <br>
                    <img src="cid:part5.09070504.00070808@psi.ch"
                      border="0"></div>
                </font></span></div>
          </blockquote>
        </div>
        <br>
      </div>
    </blockquote>
    <br>
    <br>
    <div class="moz-signature">-- <br>
      <img src="cid:part6.05070600.00070501@psi.ch" border="0"></div>
  </body>
</html>