<div dir="ltr"><div>Hi Roland,<br><br>You mentioned that I could try linking in the BLAS library from ATLAS. But running "locate libatlas.dylib" doesn't return anything. However, I do get results for libblas.dylib and liblapack.dylib.</div><div><br></div><div>I should've been clear about this: I'm trying all these things on my laptop (OS X 10.10.5), not on a cluster. So it's easier to build different versions of things if necessary.<br><br>Both my OpenBLAS and BLAS+LAPACK (Einstein Toolkit version) compilations failed. For OpenBLAS, I get the error:</div><div><br></div><div>make[3]: *** [ScalarBoundary.c.o] Error 1<br>make[2]: *** [make.checked] Error 2<br>make[1]: *** [/Users/user/Cactus/configs/ET<wbr>_OpenBLAS/lib/libthorn_Boundar<wbr>y.a] Error 2<br>make[1]: *** Waiting for unfinished jobs....<br>COMPILING configs/ET_OpenBLAS/bindings/b<wbr>uild/ADMMass/cctk_ThornBinding<wbr>s.c<br>Creating /Users/user/Cactus/configs/ET_<wbr>OpenBLAS/lib/libthorn_ADMMass.<wbr>a<br>make: *** [ET_OpenBLAS] Error 2<div><br></div>For BLAS+LAPACK, I get:<br><br>------------------------------<wbr>------------------------<br>There were 2 errors during execution of the CST<br>These must be corrected before compilation can proceed<br>------------------------------<wbr>------------------------<br><br>------------------------------<wbr>------------------------<br>Warnings were generated during execution of the CST<br>------------------------------<wbr>------------------------<br><br>CST error 1:<br> -> Configuration script for thorn BLAS returned exit code 2<br> (no error message)<br><br>CST error 2:<br> -> Configuration script for thorn LAPACK returned exit code 2<br> (no error message)<div><br>------------------------------<wbr>------------------------<br><br>make[1]: *** [/Users/user/Cactus/configs/ET<wbr>_BLAS/config-data/make.thornli<wbr>st] Error 1<br>make: *** [ET_BLAS] Error 2</div><div><br></div><div>Do you have any idea what could be going wrong?</div><div><br></div></div><div>Gwyneth</div><div><br></div></div><div class="gmail_extra"><br><div class="gmail_quote">On Mon, Mar 13, 2017 at 2:39 PM, Roland Haas <span dir="ltr"><<a href="mailto:rhaas@illinois.edu" target="_blank">rhaas@illinois.edu</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Hello Gwyneth,<br>
<span class=""><br>
> Thanks Roland! Adding -lgfortran to LIB_CXX eliminated most of the<br>
> "undefined symbols" errors.<br>
</span>Good to hear.<br>
<span class=""><br>
> The ones that remain are related to liblapack.a and liblorene_g.a. For<br>
> example:<br>
><br>
> Undefined symbols for architecture x86_64:<br>
> "_ATL_cGetNB", referenced from:<br>
> _ATL_ilaenv in liblapack.a(ATL_ilaenv.o)<br>
><br>
> I guess there's still some problem with the libraries.<br>
><br>
> Apparently one needs the development version of Lapack for a<br>
> successful Lorene compilation<br>
> (<a href="http://www.lorene.obspm.fr/prerequisites.html" rel="noreferrer" target="_blank">http://www.lorene.obspm.fr/<wbr>prerequisites.html</a>), but I think I only<br>
> have the December 2016 release. I'll see if adding the development<br>
> version makes a difference.<br>
</span>This symbol (_ATL_cGetNB) seems to be from Atlas<br>
(<a href="https://www.google.com/search?q=ATL_cGetNB" rel="noreferrer" target="_blank">https://www.google.com/<wbr>search?q=ATL_cGetNB</a>) a self-tuning LAPACK<br>
variant.<br>
<br>
There seems to be a bug in some versions of LAPACK on redhat<br>
systems that can cause this (at least this is what the google results<br>
seem to indicate). In that case you may want to contact your local<br>
cluster admin to ask about a corrected version of ATALAS/LAPACK.<br>
<br>
A workaround may be to explicitly link in the blas library from atlas<br>
usually called libatlas .<br>
<br>
If you are using the system LAPACK, you can try and see if compiling<br>
(one of the versions of) LAPACK included in the Einstein Toolkit lets<br>
you continue.<br>
<br>
There are two options: 1. use OpenBLAS which is a fast LAPACK+BLAS<br>
library 2. use the reference LAPACK and BLAS wich will almost always<br>
work but are much (factor or 10 or more) slower than optimized<br>
versions. Speed of course only matters to you if you rely on BLAS for<br>
speed (and not much in the ET does).<br>
<br>
To compile OpenBLAS please do the following:<br>
<br>
* make sure that ExternalLibraries/OpenBLAS appears in your thornlist<br>
(and is not commented out)<br>
* make sure that ExternalLibraries/BLAS and ExternalLibraries/LAPACK do<br>
*not* appear in your thorn list<br>
* set the options: OPENBLAS_DIR=BUILD in your option list and recompile<br>
<br>
If compiling OpenBLAS fails then you can also try the reference<br>
LAPACK+BLAS:<br>
<br>
* make sure that ExternalLibraries/OpenBLAS does *not* appear in your<br>
thornlist<br>
* make sure that ExternalLibraries/BLAS and ExternalLibraries/LAPACK do<br>
appear in your thorn list<br>
* set the options: BLAS_DIR=BUILD and LAPACK_DIR=BUILD in your option<br>
list and recompile<br>
<div class="HOEnZb"><div class="h5"><br>
Yours,<br>
Roland<br>
<br>
<br>
--<br>
My email is as private as my paper mail. I therefore support encrypting<br>
and signing email messages. Get my PGP key from <a href="http://pgp.mit.edu" rel="noreferrer" target="_blank">http://pgp.mit.edu</a> .<br>
Disclaimer - University of Cape Town This e-mail is subject to UCT policies and e-mail disclaimer published on our website at <a href="http://www.uct.ac.za/about/policies/emaildisclaimer/" rel="noreferrer" target="_blank">http://www.uct.ac.za/about/<wbr>policies/emaildisclaimer/</a> or obtainable from <a href="tel:%2B27%2021%20650%209111" value="+27216509111">+27 21 650 9111</a>. If this e-mail is not related to the business of UCT, it is sent by the sender in an individual capacity. Please report security incidents or abuse via <a href="mailto:csirt@uct.ac.za">csirt@uct.ac.za</a><br>
</div></div><br>______________________________<wbr>_________________<br>
Users mailing list<br>
<a href="mailto:Users@einsteintoolkit.org">Users@einsteintoolkit.org</a><br>
<a href="http://lists.einsteintoolkit.org/mailman/listinfo/users" rel="noreferrer" target="_blank">http://lists.einsteintoolkit.<wbr>org/mailman/listinfo/users</a><br>
<br></blockquote></div><br></div>