<div dir="ltr"><div><div><div><div><div>Hi Filippo,<br><br></div>I just tried the new revision and compared it with the 5.3.0 I use daily.<br></div>The name conflict can be found by grep _ESSL. I found one in PW/src/rdiaghg.f90.<br></div>I doubt this one is seriously used because it was not protected by __LINUX_ESSL for BG/Q.<br></div>I also noticed that the different implementations for ESSL and LAPACK in cdiagh.f90 v5.3.0 have been reduced in this new revision.<br><br></div>I build the code with XL C/C++ for Blue Gene/Q Fix Pack 14 (May 2016 Update) for 12.1<br>export ARCH=ppc64-bgq<br>export CC=mpixlc_r<br>export CXX=mpixlC_r<br>export F90=mpixlf90_r<br>export FC=mpixlf90_r<br><br>export BLAS_LIBS=/soft/libraries/essl<wbr>/current/essl/5.1/lib64/libess<wbr>lsmpbg.a<br>export LAPACK_LIBS="/soft/libraries/e<wbr>ssl/current/essl/5.1/lib64/lib<wbr>esslsmpbg.a /soft/libraries/alcf/current/x<wbr>l/LAPACK/lib/liblapack.a"<br>export SCALAPACK_LIBS=/soft/libraries<wbr>/alcf/current/xl/SCALAPACK/<wbr>lib/libscalapack.a<br><br><div><div>./configure --enable-openmp --with-scalapack --with-elpa<br><br></div><div>I get the code compiled by adjusting the lines 1523-1527 to use the legacy ELPA. I read your announcement that it will be deprecated. It should not be a problem.<br><br></div><div>The only big issue is that I must add the -D__OPENMP to get the preprocessing correctly handled by xlf. Without it, the results go wrong.<br></div><div>I thought __OPENMP was turned on automatically by the compiler when OpenMP was activated but clearly this was not the case for xlf.</div><div><br></div><div>Best,<br></div><div>Ye<br><br></div></div></div><div class="gmail_extra"><br clear="all"><div><div class="gmail_signature" data-smartmail="gmail_signature"><div dir="ltr">===================<br>
Ye Luo, Ph.D.<br>
Leadership Computing Facility<br>
Argonne National Laboratory</div></div></div>
<br><div class="gmail_quote">2016-09-11 16:00 GMT-05:00 Filippo SPIGA <span dir="ltr"><<a href="mailto:filippo.spiga@quantum-espresso.org" target="_blank">filippo.spiga@quantum-espresso.org</a>></span>:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Dear Ye,<br>
<span class=""><br>
On Sep 11, 2016, at 9:40 PM, Ye Luo <<a href="mailto:xw111luoye@gmail.com">xw111luoye@gmail.com</a>> wrote:<br>
> Some routines from ESSL have the same names as LAPACK but take different arguments. Also ESSL has an incomplete set of LAPACK routines.<br>
> The right order of linking involves ESSL is ESSL(for part of the LAPACK routines), LAPACK, ESSL(for BLAS).<br>
<br>
</span>Can you tell us which routines to be exact and can you test the latest snapshot?<br>
<br>
<a href="http://qe-forge.org/snapshots/espresso-r12956-2016-09-11.tar.gz" rel="noreferrer" target="_blank">http://qe-forge.org/snapshots/<wbr>espresso-r12956-2016-09-11.<wbr>tar.gz</a><br>
<br>
I will incorporate the way correct linking sequence once you can tell us it actually works without issue.<br>
<br>
Thanks in advance!<br>
<span class="im HOEnZb"><br>
--<br>
Filippo SPIGA ~ Quantum ESPRESSO Foundation ~ <a href="http://www.quantum-espresso.org" rel="noreferrer" target="_blank">http://www.quantum-espresso.<wbr>org</a><br>
<br>
<br>
</span><div class="HOEnZb"><div class="h5">______________________________<wbr>_________________<br>
Q-e-developers mailing list<br>
<a href="mailto:Q-e-developers@qe-forge.org">Q-e-developers@qe-forge.org</a><br>
<a href="http://qe-forge.org/mailman/listinfo/q-e-developers" rel="noreferrer" target="_blank">http://qe-forge.org/mailman/<wbr>listinfo/q-e-developers</a><br>
</div></div></blockquote></div><br></div>