<p>You can measure the time spent in Hypre via PCApply and PCSetUp, but you can't get finer grained integrated profiling because it was not set up that way.</p>
<div class="gmail_quote">On Sep 30, 2012 3:26 PM, "TAY wee-beng" <<a href="mailto:zonexo@gmail.com">zonexo@gmail.com</a>> wrote:<br type="attribution"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">

  
    
  
  <div bgcolor="#FFFFFF" text="#000000">
    <div>On 27/9/2012 1:44 PM, Matthew Knepley
      wrote:<br>
    </div>
    <blockquote type="cite">On Thu, Sep 27, 2012 at 3:49 AM, TAY wee-beng <span dir="ltr"><<a href="mailto:zonexo@gmail.com" target="_blank">zonexo@gmail.com</a>></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,<br>
          <br>
          I'm doing a log summary for my 3d cfd code. I have some
          questions:<br>
          <br>
          1. if I'm solving 3 linear equations using ksp, is the result
          given in the log summary the total of the 3 linear eqns'
          performance? How can I get the performance for each individual
          eqn?<br>
        </blockquote>
        <div><br>
        </div>
        <div>Use logging stages: <a href="http://www.mcs.anl.gov/petsc/petsc-dev/docs/manualpages/Profiling/PetscLogStagePush.html" target="_blank">http://www.mcs.anl.gov/petsc/petsc-dev/docs/manualpages/Profiling/PetscLogStagePush.html</a></div>

        <div> </div>
        <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
          2. If I run my code for 10 time steps, does the log summary
          gives the total or avg performance/ratio?<br>
        </blockquote>
        <div><br>
        </div>
        <div>Total.</div>
        <div> </div>
        <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
          3. Besides PETSc, I'm also using HYPRE's native geometric MG
          (Struct) to solve my Cartesian's grid CFD poisson eqn. Is
          there any way I can use PETSc's log summary to get HYPRE's
          performance? If I use boomerAMG thru PETSc, can I get its
          performance?</blockquote>
        <div><br>
        </div>
        <div>If you mean flops, only if you count them yourself and tell
          PETSc using <a href="http://www.mcs.anl.gov/petsc/petsc-dev/docs/manualpages/Profiling/PetscLogFlops.html" target="_blank">http://www.mcs.anl.gov/petsc/petsc-dev/docs/manualpages/Profiling/PetscLogFlops.html</a></div>

        <div><br>
        </div>
        <div>This is the disadvantage of using packages that do not
          properly monitor things :)</div>
        <div><br>
        </div>
        <div>    Matt</div>
        <div> </div>
      </div>
    </blockquote>
    So u mean if I use boomerAMG thru PETSc, there is no proper way of
    evaluating its performance, beside using PetscLogFlops?<br>
    <blockquote type="cite">
      <div class="gmail_quote">
        <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
          <span><font color="#888888"><br>
              -- <br>
              Yours sincerely,<br>
              <br>
              TAY wee-beng<br>
              <br>
            </font></span></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>
  </div>

</blockquote></div>