<div dir="ltr">Hypre embedded LAPACK in their source so I doubt you can have hypre use MKL. Hypre supports OMP now, but I have been having a lot of numerical problems with hypre/OMP on KNL.</div><br><div class="gmail_quote"><div dir="ltr">On Mon, Jul 30, 2018 at 9:24 AM Bastian Löhrer <<a href="mailto:bastian.loehrer@tu-dresden.de">bastian.loehrer@tu-dresden.de</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div text="#000000" bgcolor="#FFFFFF">
On 30.07.2018 13:28, Matthew Knepley wrote:<br>
<blockquote type="cite">
<div dir="ltr">
<div class="gmail_quote">
<div dir="ltr">On Mon, Jul 30, 2018 at 5:48 AM Bastian Löhrer
<<a href="mailto:bastian.loehrer@tu-dresden.de" target="_blank">bastian.loehrer@tu-dresden.de</a>>
wrote:<br>
</div>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div text="#000000" bgcolor="#FFFFFF">
<p>Dear PETSc users,</p>
<p>im configuring PETSc to use an intel stack including
intel/mkl:</p>
<tt>./configure
PETSC_ARCH=$PETSC_ARCH
\</tt><tt><br>
</tt><tt>--with-gnu-compilers=0
--with-vendor-compilers=intel
\</tt><tt><br>
</tt><tt>--with-large-file-io=1
\</tt><tt><br>
</tt><tt>--CFLAGS="-L${I_MPI_ROOT}/intel64/lib
-I${I_MPI_ROOT}/intel64/include -lmpi" \</tt><tt><br>
</tt><tt>--CXXFLAGS="-L${I_MPI_ROOT}/intel64/lib
-I${I_MPI_ROOT}/intel64/include -lmpi -lmpicxx" \</tt><tt><br>
</tt><tt>--FFLAGS="-L${I_MPI_ROOT}/intel64/lib
-I${I_MPI_ROOT}/intel64/include -lmpi" \</tt><tt><br>
</tt><tt>--LDFLAGS="-L${I_MPI_ROOT}/intel64/lib
-I${I_MPI_ROOT}/intel64/include -lmpi" \</tt><tt><br>
</tt><tt>--with-blas-lapack-dir="${MKLROOT}/lib/intel64"
\</tt><tt><br>
</tt><tt>--download-hypre
\</tt><tt><br>
</tt><tt>--with-debugging=yes</tt><br>
<br>
<b><br>
two questions:</b><br>
<br>
<b><br>
1)</b> the blas-lapack-dir option is not passed down to
the compilation of hypre according to
$PETSC_DIR/$PETSC_ARCH/conf/hypre<br>
<b>Is there a way of having PETSc compile hypre with my
intel/mkl?</b></div>
</blockquote>
<div><br>
</div>
<div>This should happen. Please send configure.log so we can
see what went on.</div>
</div>
</div>
</blockquote>
My initial guess was that the file $PETSC_DIR/$PETSC_ARCH/conf/hypre
lists the parameters which are used for compiling hypre. As I said,
this file does not mention mkl anywhere.<br>
I may be mistaken though, because having a second look I do realize
that the mkl library is mentioned in the configure log file in line
86661 ff. (where hypre is being configured) and at the end of the
log.<br>
<br>
Here is my configure.log:<br>
<a class="m_7761401067452335977moz-txt-link-freetext" href="https://cloudstore.zih.tu-dresden.de/index.php/s/b6rT0WMAKEMsj8S/download" target="_blank">https://cloudstore.zih.tu-dresden.de/index.php/s/b6rT0WMAKEMsj8S/download</a><br>
<br>
Here is that hypre file:<br>
<a class="m_7761401067452335977moz-txt-link-freetext" href="https://cloudstore.zih.tu-dresden.de/index.php/s/TSfXQ2pgDw5ALZm/download" target="_blank">https://cloudstore.zih.tu-dresden.de/index.php/s/TSfXQ2pgDw5ALZm/download</a><br>
<br>
Thanks,<br>
Bastian<br>
<br>
<br>
<blockquote type="cite">
<div dir="ltr">
<div class="gmail_quote">
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div text="#000000" bgcolor="#FFFFFF"><b><br>
2)</b> <b>Should I omit or include any option?</b> I
have come across a few options in previous configuration
calls used at my department which I have removed from my
configuration call because <br>
<blockquote>a) I had the impression that they were of no
additional use:<br>
<ul>
<li><tt>--with-cc=mpiicc --with-cxx=mpiicpc
--with-fc=mpiifort --with-mpi=1</tt></li>
<li><tt>--with-pic</tt>, because <tt>--with-pic=1</tt>
is the default anyway since 2.3.1, is it not? <br>
(<a class="m_7761401067452335977m_3080504431895625717moz-txt-link-freetext" href="https://www.mcs.anl.gov/petsc/documentation/changes/231.html" target="_blank">https://www.mcs.anl.gov/petsc/documentation/changes/231.html</a>)</li>
<li><tt>--with-mpiexec=mpirun</tt></li>
<li><tt>--with-mpi-compilers=1</tt></li>
<li><tt>--known-mpi-shared=1</tt></li>
<li><tt>--with-mpi-dir=...</tt></li>
<li><tt>COPTFLAGS="-g" CXXOPTFLAGS="-g" FOPTFLAGS="-g"</tt>
because <tt>--with-debugging=yes</tt> adds them
anyway</li>
</ul>
</blockquote>
</div>
</blockquote>
<div>Yes, getting rid of all these is fine. </div>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div text="#000000" bgcolor="#FFFFFF">
<blockquote>b) because I couldn't figure out what they
were actually for:<br>
<ul>
<li><tt><tt>--configModules=PE</tt><tt>TSc.Configure</tt></tt></li>
<li><tt><tt>--optionsModule=PETSc.compilerOptions</tt></tt><br>
</li>
</ul>
</blockquote>
<blockquote> </blockquote>
</div>
</blockquote>
<div>Those are added automatically. These are hooks so that
you can completely change the system without</div>
<div>getting rid of the low-level tests. </div>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div text="#000000" bgcolor="#FFFFFF">
<blockquote>
<p>c) others:</p>
<ul>
<li><tt>--known-64-bit-blas-indices</tt> I guess it
wouldn't hurt anyway so I guess I'll include this
option the next time I configure petsc<br>
(<a class="m_7761401067452335977m_3080504431895625717moz-txt-link-freetext" href="http://www.mcs.anl.gov/petsc/petsc-current/docs/manualpages/Sys/PetscBLASInt.html" target="_blank">http://www.mcs.anl.gov/petsc/petsc-current/docs/manualpages/Sys/PetscBLASInt.html</a>)<br>
</li>
</ul>
</blockquote>
</div>
</blockquote>
<div>We have a bunch of -known-* options. They are used for
any test that has to execute. Batch environments cannot,</div>
<div>and thus you need to specify all of them. You do not need
it unless you are using --with-batch.</div>
</div>
</div>
</blockquote>
Okay, fantastic, thank you!<br>
</div>
</blockquote></div>