Compiling issue.

From NWChem

Viewed 1465 times, With a total of 19 Posts
Jump to: navigation, search

Jump to page 12Next 16Last
Clicked A Few Times
Threads 2
Posts 12
Dear All.

I have been trying to install NWChem 6.5 with the follow environmental variables and so far I have not had succeed.

setenv NWCHEM_TOP /home/frank/bin/nwchem-6.5
setenv NWCHEM_TARGET LINUX64
setenv NWCHEM_MODULES "all"
setenv NWCHEM_LONG_PATHS Y
setenv USE_NOFSCHECK Y
setenv USE_MPI y
setenv USE_MPIF y
setenv USE_MPIF4 y
setenv MPI_INCLUDE "/home/frank/bin/openmpi/include"
setenv MPI_LIB "/home/frank/bin/openmpi/lib"
setenv LIBMPI " -lmpi_usempi -lmpi_mpifh -lmpi"
setenv BLASOPT="-L/home/frank/bin/acml/gfortran64/lib -lacml"
setenv CC gcc
setenv CXX g++
setenv FC gfortran
cd $NWCHEM_TOP/src
make nwchem_config NWCHEM_MODULES=all
make

When I followed the progress on screen, the make procedure was performing a lot of process, see next, after one hour (roughly) I killed the job since it was taking so much time, usually the compilations on my PC take about 30 min.

If anyone can help me to overcome this problem I would be very grateful.

Kind regards, Frank


make[80]: *** [clean] Interrupt
make[79]: *** [clean] Interrupt
make[78]: *** [clean] Interrupt
make[77]: *** [build/config.status] Interrupt
make[76]: *** [libraries] Interrupt
make[75]: *** [build/config.status] Interrupt
make[74]: *** [libraries] Interrupt
make[73]: *** [build/config.status] Interrupt
make[72]: *** [libraries] Interrupt
make[71]: *** [build/config.status] Interrupt
make[70]: *** [libraries] Interrupt
make[69]: *** [build/config.status] Interrupt
make[68]: *** [libraries] Interrupt
make[67]: *** [build/config.status] Interrupt
make[66]: *** [libraries] Interrupt
make[65]: *** [build/config.status] Interrupt
make[64]: *** [libraries] Interrupt
make[63]: *** [build/config.status] Interrupt
make[62]: *** [libraries] Interrupt
make[61]: *** [build/config.status] Interrupt
make[60]: *** [libraries] Interrupt
make[59]: *** [build/config.status] Interrupt
make[58]: *** [libraries] Interrupt
make[57]: *** [build/config.status] Interrupt
make[56]: *** [libraries] Interrupt
make[55]: *** [build/config.status] Interrupt
make[54]: *** [libraries] Interrupt
make[53]: *** [build/config.status] Interrupt
make[52]: *** [libraries] Interrupt
make[51]: *** [build/config.status] Interrupt
make[50]: *** [libraries] Interrupt
make[49]: *** [build/config.status] Interrupt
make[48]: *** [libraries] Interrupt
make[47]: *** [build/config.status] Interrupt
make[46]: *** [libraries] Interrupt
make[45]: *** [build/config.status] Interrupt
make[44]: *** [libraries] Interrupt
make[43]: *** [build/config.status] Interrupt
make[42]: *** [libraries] Interrupt
make[41]: *** [build/config.status] Interrupt
make[40]: *** [libraries] Interrupt
make[39]: *** [build/config.status] Interrupt
make[38]: *** [libraries] Interrupt
make[37]: *** [build/config.status] Interrupt
make[36]: *** [libraries] Interrupt
make[35]: *** [build/config.status] Interrupt
make[34]: *** [libraries] Interrupt
make[33]: *** [build/config.status] Interrupt
make[32]: *** [libraries] Interrupt
make[31]: *** [build/config.status] Interrupt
make[30]: *** [libraries] Interrupt
make[29]: *** [build/config.status] Interrupt
make[28]: *** [libraries] Interrupt
make[27]: *** [build/config.status] Interrupt
make[26]: *** [libraries] Interrupt
make[25]: *** [build/config.status] Interrupt
make[24]: *** [libraries] Interrupt
make[23]: *** [build/config.status] Interrupt
make[22]: *** [libraries] Interrupt
make[21]: *** [build/config.status] Interrupt
make[20]: *** [libraries] Interrupt
make[19]: *** [build/config.status] Interrupt
make[18]: *** [libraries] Interrupt
make[17]: *** [build/config.status] Interrupt
make[16]: *** [libraries] Interrupt
make[15]: *** [build/config.status] Interrupt
make[14]: *** [libraries] Interrupt
make[13]: *** [build/config.status] Interrupt
make[12]: *** [libraries] Interrupt
make[11]: *** [build/config.status] Interrupt
make[10]: *** [libraries] Interrupt
make[9]: *** [build/config.status] Interrupt
make[8]: *** [libraries] Interrupt
make[7]: *** [build/config.status] Interrupt
make[6]: *** [libraries] Interrupt
make[5]: *** [build/config.status] Interrupt
make[4]: *** [libraries] Interrupt
make[3]: *** [build/config.status] Interrupt
make[2]: *** [libraries] Interrupt
make[1]: *** [build/config.status] Interrupt
make: *** [libraries] Interrupt

Forum Vet
Threads 4
Posts 936
Frank
Could you restart the compilation by using the following command and posting the complete compilation log?
Thanks


make -j1 -l 0.00001

Clicked A Few Times
Threads 2
Posts 12
Hi Edo.

This is part of the output. I did not wait until the compilation finish because It was behaving pretty much as the old one.

      • Configuring Parallel Tools ****

../ga-5-3/configure --prefix=/home/frank/bin/nwchem-6.5/src/tools/install --with-tcgmsg --with-mpi=-I/home/frank/bin/openmpi/include -L/home/frank/bin/openmpi/lib -lmpi_f90 -lmpi_f77 -lmpi -ldl -lm -Wl,--export-dynamic -lrt -lnsl -lutil -lm -ldl /home/frank/bin/openmpi --enable-peigs --enable-underscoring --disable-mpi-tests --without-scalapack --without-lapack --with-blas8=-L//opt/acml-6/gfortran64/lib/ -lacml --with-sockets CC=cc F77=gfortran FFLAGS=-fno-aggressive-loop-optimizations ARMCI_DEFAULT_SHMMAX_UBOUND=131072
checking build system type... x86_64-unknown-linux-gnu
checking host system type... x86_64-unknown-linux-gnu
checking for TARGET base (64bit-ness checked later)... LINUX
checking whether we think this system is what we call SYSV... yes
checking whether we are cross compiling... no
checking for a BSD-compatible install... cat: /home/frank/bin/nwchem-6.5/src/tools/build/config.h: No such file or directory
make[4]: Entering directory `/home/frank/bin/nwchem-6.5/src'
test \! -f 64_to_32 -o \! -f 32_to_64 || rm -f 64_to_32 32_to_64
test -d /home/frank/bin/nwchem-6.5/lib/LINUX64 || mkdir -p /home/frank/bin/nwchem-6.5/lib/LINUX64
test -d /home/frank/bin/nwchem-6.5/bin/LINUX64 || mkdir -p /home/frank/bin/nwchem-6.5/bin/LINUX64
./util/util_nwchem_version.bash
Making libraries in tools
cat: /home/frank/bin/nwchem-6.5/src/tools/build/config.h: No such file or directory
make[5]: warning: -jN forced in submake: disabling jobserver mode.

      • Configuring Parallel Tools ****

../ga-5-3/configure --prefix=/home/frank/bin/nwchem-6.5/src/tools/install --with-tcgmsg --with-mpi=-I/home/frank/bin/openmpi/include -L/home/frank/bin/openmpi/lib -lmpi_f90 -lmpi_f77 -lmpi -ldl -lm -Wl,--export-dynamic -lrt -lnsl -lutil -lm -ldl /home/frank/bin/openmpi --enable-peigs --enable-underscoring --disable-mpi-tests --without-scalapack --without-lapack --with-blas8=-L//opt/acml-6/gfortran64/lib/ -lacml --with-sockets CC=cc F77=gfortran FFLAGS=-fno-aggressive-loop-optimizations ARMCI_DEFAULT_SHMMAX_UBOUND=131072
checking build system type... x86_64-unknown-linux-gnu
checking host system type... x86_64-unknown-linux-gnu
checking for TARGET base (64bit-ness checked later)... LINUX
checking whether we think this system is what we call SYSV... yes
checking whether we are cross compiling... no
checking for a BSD-compatible install... cat: /home/frank/bin/nwchem-6.5/src/tools/build/config.h: No such file or directory
make[6]: Entering directory `/home/frank/bin/nwchem-6.5/src'
test \! -f 64_to_32 -o \! -f 32_to_64 || rm -f 64_to_32 32_to_64
test -d /home/frank/bin/nwchem-6.5/lib/LINUX64 || mkdir -p /home/frank/bin/nwchem-6.5/lib/LINUX64
test -d /home/frank/bin/nwchem-6.5/bin/LINUX64 || mkdir -p /home/frank/bin/nwchem-6.5/bin/LINUX64
./util/util_nwchem_version.bash
Making libraries in tools
cat: /home/frank/bin/nwchem-6.5/src/tools/build/config.h: No such file or directory
make[7]: warning: -jN forced in submake: disabling jobserver mode.

      • Configuring Parallel Tools ****

../ga-5-3/configure --prefix=/home/frank/bin/nwchem-6.5/src/tools/install --with-tcgmsg --with-mpi=-I/home/frank/bin/openmpi/include -L/home/frank/bin/openmpi/lib -lmpi_f90 -lmpi_f77 -lmpi -ldl -lm -Wl,--export-dynamic -lrt -lnsl -lutil -lm -ldl /home/frank/bin/openmpi --enable-peigs --enable-underscoring --disable-mpi-tests --without-scalapack --without-lapack --with-blas8=-L//opt/acml-6/gfortran64/lib/ -lacml --with-sockets CC=cc F77=gfortran FFLAGS=-fno-aggressive-loop-optimizations ARMCI_DEFAULT_SHMMAX_UBOUND=131072
^Cmake[7]: *** [build/config.status] Interrupt
make[6]: *** [libraries] Interrupt
make[5]: *** [build/config.status] Interrupt
make[4]: *** [libraries] Interrupt
make[3]: *** [build/config.status] Interrupt
make[2]: *** [libraries] Interrupt
make[1]: *** [build/config.status] Interrupt
make: *** [libraries] Interrupt

Forum Vet
Threads 4
Posts 936
Could you send (or upload to a accessible website) the following file
$NWCHEM_TOP/src/tools/build/config.log

Clicked A Few Times
Threads 2
Posts 12
Hi Edo.

Feel free to download the file from the next link.

https://www.dropbox.com/s/w9i1izje5eqqtc6/config.log?dl=0

Thanks

Forum Vet
Threads 4
Posts 936
Frank
Something goes wrong when configure look for your install binary

Please send me the output of the following commands

which install
install -c
/usr/bin/install -c

Clicked A Few Times
Threads 2
Posts 12
Hi Edo.

I named my installation script as "install" and it was located in my $home/bin so, when I was trying to compile the program, the make process was not calling the system install located in /usr/bin but the one located in my $home/bin/. I corrected that and the NWChem installation was completely fine, but, when I execute the program it runs OK on a single core but the parallel does not, this is the tail of the .out file.


The calculation does not go any further of the last line " Singularity in Pulay matrix. Error and Fock matrices removed"

(Hope I made myself clear)

Thank you so much.

Grid_pts file          = ./scan.tmp.gridpts.0
Record size in doubles = 12289 No. of grid_pts per rec = 3070
Max. records in memory = 4 Max. recs in file = 231234


          Memory utilization after 1st SCF pass: 
Heap Space remaining (MW): 65.49 65486212
Stack Space remaining (MW): 65.54 65535716

  convergence    iter        energy       DeltaE   RMS-Dens  Diis-err    time
---------------- ----- ----------------- --------- --------- --------- ------
d= 0,ls=0.0,diis 1 -14.9106132446 -1.56D+01 9.98D-03 5.97D-03 0.2
Singularity in Pulay matrix. Error and Fock matrices removed.

Forum Vet
Threads 4
Posts 936
I am not 100% sure if the following is causing your problems when running in parallel,
but it's for sure likely to cause problems.

setenv BLASOPT="-L/home/frank/bin/acml/gfortran64/lib -lacml"

should be changed to

setenv BLASOPT="-L/home/frank/bin/acml/gfortran64_int64/lib -lacml"

More details at
http://www.nwchem-sw.org/index.php/Compiling_NWChem#Optimized_math_libraries

Clicked A Few Times
Threads 2
Posts 12
Dear Edo.

I changed my environmental variables this time. I though that the problem was related to the ACML library version that I was using (6) so I compiled the code again but with the 5.3.1 version. (These are my new environmental variables)

  1. !/bin/bash
export NWCHEM_TOP=/home/frank/bin/nwchem-6.5
export LARGE_FILES=TRUE
export NWCHEM_TARGET=LINUX64
export NWCHEM_MODULES="all"
export USE_MPI=y
export USE_MPIF=y
export USE_MPIF4=y
export MPI_LOC=/home/frank/bin/openmpi
export MPI_LIB=/home/frank/bin/openmpi/lib
export MPI_INCLUDE=/home/frank/bin/openmpi/include
export LIBMPI="-lmpi_usempi -lmpi_mpifh -lmpi"
  1. export FC=gfortran
  2. export CC=gcc
  3. export CXX=g++
export ARMCI_NETWORK=SOCKETS
export ACMLLIB=/opt/acml5.3.1/gfortran64_int64/lib
export BLASOPT="-L/$ACMLLIB/ -lacml"

cd $NWCHEM_TOP/src

make realclean

make nwchem_config
make -j4

cd $NWCHEM_TOP/contrib
setenv FC ifort

./getmem.nwchem

The same problem appears,

  convergence    iter        energy       DeltaE   RMS-Dens  Diis-err    time
---------------- ----- ----------------- --------- --------- --------- ------
d= 0,ls=0.0,diis 1 -14.9106132446 -1.56D+01 9.98D-03 5.97D-03 0.2
Singularity in Pulay matrix. Error and Fock matrices removed.

Thanks!

Clicked A Few Times
Threads 2
Posts 12
I noticed something about the NWChem calculations, when I tried to run a DFT calculation the error that came out is the last that I posted, but, when the calculation is an Ab-initio the error is different

ga_iter_lsolve: dgesv failed 140733193388032

It happens either single or multicore.

Hope this information worth.

Forum Vet
Threads 4
Posts 936
Quote:Frank.ramirez Mar 19th 11:02 am
I noticed something about the NWChem calculations, when I tried to run a DFT calculation the error that came out is the last that I posted, but, when the calculation is an Ab-initio the error is different

ga_iter_lsolve: dgesv failed 140733193388032

It happens either single or multicore.

Hope this information worth.

This is a clear indication of a BLASOPT problem. Could you please send us the complete output of the commands


ldd $NWCHEM_TOP/bin/LINUX64/nwchem
cd $NWCHEM_TOP/src; make FC=gfortran link

Clicked A Few Times
Threads 2
Posts 12
linux-vdso.so.1 => (0x00007fffdbd18000)
libacml.so => /opt/acml5.3.1/gfortran64_int64/lib/libacml.so (0x00002afd8b968000)
libmpi_usempi.so.1 => /home/frank/bin/openmpi/lib/libmpi_usempi.so.1 (0x00002afd8dd98000)
libmpi_mpifh.so.2 => /home/frank/bin/openmpi/lib/libmpi_mpifh.so.2 (0x00002afd8dfa0000)
libmpi.so.1 => /home/frank/bin/openmpi/lib/libmpi.so.1 (0x00002afd8e1f8000)
libgfortran.so.3 => /usr/lib/x86_64-linux-gnu/libgfortran.so.3 (0x00002afd8e4d8000)
libm.so.6 => /lib/x86_64-linux-gnu/libm.so.6 (0x00002afd8e7f8000)
libc.so.6 => /lib/x86_64-linux-gnu/libc.so.6 (0x00002afd8eb00000)
librt.so.1 => /lib/x86_64-linux-gnu/librt.so.1 (0x00002afd8eec8000)
libopen-pal.so.6 => /home/frank/bin/openmpi/lib/libopen-pal.so.6 (0x00002afd8f0d0000)
libpthread.so.0 => /lib/x86_64-linux-gnu/libpthread.so.0 (0x00002afd8f3b8000)
libopen-rte.so.7 => /home/frank/bin/openmpi/lib/libopen-rte.so.7 (0x00002afd8f5d8000)
libquadmath.so.0 => /usr/lib/x86_64-linux-gnu/libquadmath.so.0 (0x00002afd8f858000)
libgcc_s.so.1 => /lib/x86_64-linux-gnu/libgcc_s.so.1 (0x00002afd8fa98000)
/lib64/ld-linux-x86-64.so.2 (0x00002afd8b740000)
libdl.so.2 => /lib/x86_64-linux-gnu/libdl.so.2 (0x00002afd8fcb0000)
libutil.so.1 => /lib/x86_64-linux-gnu/libutil.so.1 (0x00002afd8feb8000)


make nwchem.o stubs.o
make[1]: warning: -jN forced in submake: disabling jobserver mode.
gfortran -m64 -ffast-math -Warray-bounds -fdefault-integer-8 -O2 -g -fno-aggressive-loop-optimizations -g -O -I. -I/home/frank/bin/nwchem-6.5/src/include -I/home/frank/bin/nwchem-6.5/src/tools/install/include -DGFORTRAN -DCHKUNDFLW -DGCC4 -DGCC46 -DEXT_INT -DLINUX -DLINUX64 -DPARALLEL_DIAG -DCOMPILATION_DATE="'`date +%a_%b_%d_%H:%M:%S_%Y`'" -DCOMPILATION_DIR="'/home/frank/bin/nwchem-6.5'" -DNWCHEM_BRANCH="'Development'" -c -o nwchem.o nwchem.F
gfortran -m64 -ffast-math -Warray-bounds -fdefault-integer-8 -O2 -g -fno-aggressive-loop-optimizations -g -O -I. -I/home/frank/bin/nwchem-6.5/src/include -I/home/frank/bin/nwchem-6.5/src/tools/install/include -DGFORTRAN -DCHKUNDFLW -DGCC4 -DGCC46 -DEXT_INT -DLINUX -DLINUX64 -DPARALLEL_DIAG -DCOMPILATION_DATE="'`date +%a_%b_%d_%H:%M:%S_%Y`'" -DCOMPILATION_DIR="'/home/frank/bin/nwchem-6.5'" -DNWCHEM_BRANCH="'Development'" -c -o stubs.o stubs.F
gfortran -L/home/frank/bin/nwchem-6.5/lib/LINUX64 -L/home/frank/bin/nwchem-6.5/src/tools/install/lib -o /home/frank/bin/nwchem-6.5/bin/LINUX64/nwchem nwchem.o stubs.o -lnwctask -lccsd -lmcscf -lselci -lmp2 -lmoints -lstepper -ldriver -loptim -lnwdft -lgradients -lcphf -lesp -lddscf -ldangchang -lguess -lhessian -lvib -lnwcutil -lrimp2 -lproperty -lsolvation -lnwints -lprepar -lnwmd -lnwpw -lofpw -lpaw -lpspw -lband -lnwpwlib -lcafe -lspace -lanalyze -lqhop -lpfft -ldplot -ldrdy -lvscf -lqmmm -lqmd -letrans -lpspw -ltce -lbq -lmm -lcons -lperfm -ldntmc -lccca -lnwcutil -lga -larmci -lpeigs -lperfm -lcons -lbq -lnwcutil -L//opt/acml5.3.1/gfortran64_int64/lib/ -lacml -llapack -lblas -L/home/frank/bin/openmpi/lib -lmpi_usempi -lmpi_mpifh -lmpi

/usr/bin/ld: Warning: alignment 16 of symbol `cface_' in /home/frank/bin/nwchem-6.5/lib/LINUX64/libstepper.a(stpr_face.o) is smaller than 32 in /home/frank/bin/nwchem-6.5/lib/LINUX64/libstepper.a(stpr_partit.o)


Thanks Edo!

Forum Vet
Threads 4
Posts 936
The output of ldd looks good.
Are you sure that this reflects what happens during your runs?

Clicked A Few Times
Threads 1
Posts 7
Similar problems
I am having a similar problem with Nwchem-6.5.revision26243-src.2014-09-10.tar.gz with OpenMPI 1.6.5 and the intel 13.0.2 compiler and intel mkl libraries.

SCF calculations are just fine, but DFT jobs have the Singularity in the Pulay Matrix error.

Did Frank find an answer?

Forum Vet
Threads 4
Posts 936
Matthew
Could you post the first 10 lines of $NWCHEM_TOP/src/tools/build/config.log


Forum >> NWChem's corner >> Compiling NWChem
Jump to page 12Next 16Last



Who's here now Members 0 Guests 1 Bots/Crawler 0


AWC's: 2.5.10 MediaWiki - Stand Alone Forum Extension
Forum theme style by: AWC