<div dir="ltr"><div><div><div><div>Dear Mr. Filippo,<br><br></div>When I run the command : make -f Makefile.gpu pw-gpu, it gives the following errors. The version of cc I use is cc (GCC) 4.4.7 20120313 (Red Hat 4.4.7-4). <br>
<br>cc -O3 -fopenmp -fPIC -z muldefs -o libphigemm.so.2.0 phigemm_auxiliary.o phigemm_env.o phigemm_dgemm.o phigemm_zgemm.o phigemm_dgemm_specialK.o phigemm_zgemm_specialK.o phigemm_cgemm.o phigemm_sgemm.o -L/usr/local/cuda/lib64 -lcublas -lcufft -lcudart /home/wang_lab/espresso-5.0.2/GPU/..//phiGEMM/lib/libphigemm.a /home/wang_lab/espresso-5.0.2/GPU/..//BLAS/blas.a -L/usr/local/cuda/lib64 -lcublas -lcufft -lcudart <br>
/usr/lib/gcc/x86_64-redhat-linux/4.4.7/../../../../lib64/crt1.o: In function `_start':<br>(.text+0x20): undefined reference to `main'<br>/home/wang_lab/espresso-5.0.2/GPU/..//BLAS/blas.a(xerbla.o): In function `xerbla_':<br>
xerbla.f:(.text+0x51): undefined reference to `_gfortran_st_write'<br>xerbla.f:(.text+0x61): undefined reference to `_gfortran_transfer_character'<br>xerbla.f:(.text+0x71): undefined reference to `_gfortran_transfer_integer'<br>
xerbla.f:(.text+0x79): undefined reference to `_gfortran_st_write_done'<br>xerbla.f:(.text+0x83): undefined reference to `_gfortran_stop_numeric'<br>collect2: ld returned 1 exit status<br>make[3]: *** [shared] Error 1<br>
make[3]: Leaving directory `/home/wang_lab/espresso-5.0.2/phiGEMM/src'<br>make[2]: *** [phigemm] Error 2<br>make[2]: Leaving directory `/home/wang_lab/espresso-5.0.2/phiGEMM'<br>make[1]: *** [libphiGEMM] Error 2<br>
make[1]: Leaving directory `/home/wang_lab/espresso-5.0.2/GPU/install'<br>make: *** [libphiGEMM] Error 2<br><br></div>Is it possible the version of the cc cause the problem? Which version you recommend? Thanks a lot for your help.<br>
<br></div>Best,<br></div>Lijuan<br><div><div><div><div><div><div><br></div></div></div></div></div></div></div><div class="gmail_extra"><br><br><div class="gmail_quote">2014-06-13 12:44 GMT-04:00 Filippo Spiga <span dir="ltr"><<a href="mailto:spiga.filippo@gmail.com" target="_blank">spiga.filippo@gmail.com</a>></span>:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Dear Lijuan,<br>
<br>
always the latest if possible. At this stage better wait for a formal supported tag for QE 5.1. It is still a good practice before trying QE-GPU to try the latest QE running CPU only.<br>
<br>
Regards,<br>
F<br>
<div class="HOEnZb"><div class="h5"><br>
On Jun 13, 2014, at 4:42 PM, Lijuan He <<a href="mailto:helijuan130@gmail.com">helijuan130@gmail.com</a>> wrote:<br>
<br>
> Dear Mr. Filippo,<br>
><br>
> Thank you so much for your help. I am wondering if I move to espresso 5.1, which version of qe-gpu I should use? Is it the QE-GPU-14.03.0.tar.gz or QE-GPU-14.01.0.tar.gz? Thanks.<br>
><br>
> Best,<br>
> Lijuan<br>
> PHD in Gatech<br>
><br>
><br>
> 2014-06-13 11:32 GMT-04:00 Filippo Spiga <<a href="mailto:spiga.filippo@gmail.com">spiga.filippo@gmail.com</a>>:<br>
> Dear Lijuan,<br>
><br>
> the old QE-GPU version has a old a different configure syntax. If you do "./configure --help" you will notice the differences. In your specific case, try "--enable-phigemm --enable-magma" and "--with-gpu-arch=35"<br>
><br>
> Anyway, just an advice: unless there is specific reason you need to use 5.0.3, I strongly suggest to move to QE 5.0.x or 5.1<br>
><br>
> Regards,<br>
> Filippo<br>
><br>
><br>
> On Jun 12, 2014, at 9:16 PM, Lijuan He <<a href="mailto:helijuan130@gmail.com">helijuan130@gmail.com</a>> wrote:<br>
> > Dear Mr. Filippo,<br>
> ><br>
> > Thanks. When I run the which nvcc, it shows : /usr/local/cuda/bin/nvcc<br>
> ><br>
> > The version I use is espresso-5.0.2 and QE-GPU-r216.tar.gz.<br>
> ><br>
> > When I tried to run the configure command: ./configure -disable-parallel --enable-openmp \--enable-cuda --with-gpu-arch=sm_35 \--with-cuda-dir=/usr/local/cuda \--with-magma --with-phigemm , the following error came up:<br>
> ><br>
> > configure: WARNING: unrecognized options: --with-magma, --with-phigemm<br>
> ><br>
> > However, at the end, it shows the configure is success. I checked the intermediate steps, it shows 'checking for nvcc... no'<br>
> ><br>
> > And then when I tried to run the command :<br>
> ><br>
> > cd ..<br>
> > make -f Makefile.gpu pw-gpu<br>
> ><br>
> > at the end, it shows two errors:<br>
> ><br>
> > mkdir -p ../bin ../lib<br>
> > ar ruv libphigemm.a phigemm_auxiliary.o phigemm_env.o phigemm_dgemm.o phigemm_zgemm.o phigemm_dgemm_specialK.o phigemm_zgemm_specialK.o phigemm_cgemm.o phigemm_sgemm.o<br>
> > ar: creating libphigemm.a<br>
> > ar: phigemm_auxiliary.o: No such file or directory<br>
> > make[3]: *** [static] Error 1<br>
> > make[3]: Leaving directory `/usr/local/espresso-5.0.2/phiGEMM/src'<br>
> > make[2]: *** [phigemm] Error 2<br>
> > make[2]: Leaving directory `/usr/local/espresso-5.0.2/phiGEMM'<br>
> > make[1]: *** [libphiGEMM] Error 2<br>
> > make[1]: Leaving directory `/usr/local/espresso-5.0.2/GPU/install'<br>
> > make: *** [libphiGEMM] Error 2<br>
> ><br>
> > Thanks for your help.<br>
> ><br>
> > Best,<br>
> > Lijuan He<br>
> > PHD in Gatech<br>
> ><br>
> ><br>
> ><br>
> ><br>
> ><br>
> ><br>
> ><br>
> > 2014-06-07 9:57 GMT-04:00 Filippo Spiga <<a href="mailto:spiga.filippo@gmail.com">spiga.filippo@gmail.com</a>>:<br>
> > Dead Lijuan,<br>
> ><br>
> > is nvcc in PATH? What happen if you run<br>
> ><br>
> > $ which nvcc<br>
> ><br>
> > ?<br>
> ><br>
> > On Jun 7, 2014, at 4:39 AM, Lijuan He <<a href="mailto:helijuan130@gmail.com">helijuan130@gmail.com</a>> wrote:<br>
> ><br>
> > > Dear all,<br>
> > ><br>
> > > I am trying to run qe on GPU. I follow the step posted on <a href="https://github.com/fspiga/QE-GPU" target="_blank">https://github.com/fspiga/QE-GPU</a> . However, when I try to run the command:<br>
> > > $ ./configure --disable-parallel --enable-openmp \<br>
> > > --enable-cuda --with-gpu-arch=sm_35 \<br>
> > > --with-cuda-dir=<full-path-where-CUDA-is-installed> \<br>
> > > --with-magma --with-phigemm<br>
> > ><br>
> > ><br>
> > > it shows that the nvcc cannot be found. But I checked the file nvcc is there. Anyone met the same problem or know what could be the possible reason? Thanks a lot.<br>
> > ><br>
> > > Best,<br>
> > > Lijuan He<br>
> > > PHD Candidate in Georgia Tech<br>
> > ><br>
> > ><br>
> > > _______________________________________________<br>
> > > Pw_forum mailing list<br>
> > > <a href="mailto:Pw_forum@pwscf.org">Pw_forum@pwscf.org</a><br>
> > > <a href="http://pwscf.org/mailman/listinfo/pw_forum" target="_blank">http://pwscf.org/mailman/listinfo/pw_forum</a><br>
> ><br>
> > --<br>
> > Mr. Filippo SPIGA, M.Sc.<br>
> > <a href="http://www.linkedin.com/in/filippospiga" target="_blank">http://www.linkedin.com/in/filippospiga</a> ~ skype: filippo.spiga<br>
> ><br>
> > «Nobody will drive us out of Cantor's paradise.» ~ David Hilbert<br>
> ><br>
> > *****<br>
> > Disclaimer: "Please note this message and any attachments are CONFIDENTIAL and may be privileged or otherwise protected from disclosure. The contents are not to be disclosed to anyone other than the addressee. Unauthorized recipients are requested to preserve this confidentiality and to advise the sender immediately of any error in transmission."<br>
> ><br>
> ><br>
> ><br>
> > _______________________________________________<br>
> > Pw_forum mailing list<br>
> > <a href="mailto:Pw_forum@pwscf.org">Pw_forum@pwscf.org</a><br>
> > <a href="http://pwscf.org/mailman/listinfo/pw_forum" target="_blank">http://pwscf.org/mailman/listinfo/pw_forum</a><br>
> ><br>
> > _______________________________________________<br>
> > Pw_forum mailing list<br>
> > <a href="mailto:Pw_forum@pwscf.org">Pw_forum@pwscf.org</a><br>
> > <a href="http://pwscf.org/mailman/listinfo/pw_forum" target="_blank">http://pwscf.org/mailman/listinfo/pw_forum</a><br>
><br>
> --<br>
> Mr. Filippo SPIGA, M.Sc.<br>
> <a href="http://www.linkedin.com/in/filippospiga" target="_blank">http://www.linkedin.com/in/filippospiga</a> ~ skype: filippo.spiga<br>
><br>
> «Nobody will drive us out of Cantor's paradise.» ~ David Hilbert<br>
><br>
> *****<br>
> Disclaimer: "Please note this message and any attachments are CONFIDENTIAL and may be privileged or otherwise protected from disclosure. The contents are not to be disclosed to anyone other than the addressee. Unauthorized recipients are requested to preserve this confidentiality and to advise the sender immediately of any error in transmission."<br>
><br>
><br>
><br>
> _______________________________________________<br>
> Pw_forum mailing list<br>
> <a href="mailto:Pw_forum@pwscf.org">Pw_forum@pwscf.org</a><br>
> <a href="http://pwscf.org/mailman/listinfo/pw_forum" target="_blank">http://pwscf.org/mailman/listinfo/pw_forum</a><br>
><br>
> _______________________________________________<br>
> Pw_forum mailing list<br>
> <a href="mailto:Pw_forum@pwscf.org">Pw_forum@pwscf.org</a><br>
> <a href="http://pwscf.org/mailman/listinfo/pw_forum" target="_blank">http://pwscf.org/mailman/listinfo/pw_forum</a><br>
<br>
--<br>
Mr. Filippo SPIGA, M.Sc.<br>
<a href="http://www.linkedin.com/in/filippospiga" target="_blank">http://www.linkedin.com/in/filippospiga</a> ~ skype: filippo.spiga<br>
<br>
«Nobody will drive us out of Cantor's paradise.» ~ David Hilbert<br>
<br>
*****<br>
Disclaimer: "Please note this message and any attachments are CONFIDENTIAL and may be privileged or otherwise protected from disclosure. The contents are not to be disclosed to anyone other than the addressee. Unauthorized recipients are requested to preserve this confidentiality and to advise the sender immediately of any error in transmission."<br>
<br>
<br>
<br>
_______________________________________________<br>
Pw_forum mailing list<br>
<a href="mailto:Pw_forum@pwscf.org">Pw_forum@pwscf.org</a><br>
<a href="http://pwscf.org/mailman/listinfo/pw_forum" target="_blank">http://pwscf.org/mailman/listinfo/pw_forum</a><br>
</div></div></blockquote></div><br></div>