Dear Vlad,
Thanks for letting us know. I tested in my local system with gcc 8.2.1, and
there were no issues, but I was using boost 1.54. Boost is quite variable
depending on the OS/version you are using, and they have been changing the
way they handle the multithreading (we had some issues with deprecated mt
libraries already). Could you check what version you have in your system? A
solution that I considered before is to include a stripped down version of
boost with the minimal library distribution, so that is not so dependent on
the system where it is being installed...we might reconsider that option.
Anyway, thanks again. If you use the code or test it, I'm always looking
forward for feedback :D.
Cheers,
El mié., 7 nov. 2018 a las 16:55, Vlad Cojocaru (<
vlad.cojocaru.mpi-muenster.mpg.de>) escribió:
> Dear Dave, Dear Stephan,
>
> Thanks for answering. So, I just managed a few minutes ago to solve the
> issue by adding the "-lpthread" explicitly to LIBS in the Makefiles of
> packmol-memgen ... (both src and python). I modified all Makefiles and I
> don't know if all modifications are needed but with them, the serial
> compilation worked now ....
>
> Anyhow, maybe this helps to add a patch that fixes this issue ...
>
> Best wishes
> Vlad
>
>
>
> On 11/7/18 4:21 PM, Stephan Schott wrote:
> > Hi Vlad,
> > I have not tried with that particular version of OS/GCC. It seems like
> some
> > library links might have changed; I will do some tests and get back to
> you.
> > In the meanwhile, after you configure AMBER, you can go to the config.h
> > file that is created, and change the very last line from MEMEMBED=yes to
> > MEMEMBED=no, and the compiling process should continue. Let me know if it
> > works.
> > Cheers,
> >
> > El mié., 7 nov. 2018 a las 14:05, David A Case (<david.case.rutgers.edu
> >)
> > escribió:
> >
> >> On Tue, Nov 06, 2018, Vlad Cojocaru wrote:
> >>> g++ -Wall -Wextra -Werror -O3 -fPIC -std=c++11 -Isrc -I../boost/1.66/
> >>> src/main.o src/pdb.o src/ga.o src/grid.o src/direct.o
> >>>
> >>
> -Wl,-rpath=/apps/amber/18_tools-18_openmpi-3.1.3_gnu-8.2.1/AmberTools/src/packmol_memgen/packmol_memgen/lib/memembed/../boost/1.66/stage/lib/
> >>
> >>
> -L/apps/amber/18_tools-18_openmpi-3.1.3_gnu-8.2.1/AmberTools/src/packmol_memgen/packmol_memgen/lib/memembed/../boost/1.66/stage/lib/
> >>
> >>> -lboost_thread -lboost_system -o bin/memembed
> >>>
> /usr/lib64/gcc/x86_64-suse-linux/8/../../../../x86_64-suse-linux/bin/ld:
> >>> src/main.o: undefined reference to symbol
> >>> 'pthread_condattr_setclock..GLIBC_2.3.3'
> >>>
> /usr/lib64/gcc/x86_64-suse-linux/8/../../../../x86_64-suse-linux/bin/ld:
> >>> /lib64/libpthread.so.0: error adding symbols: DSO missing from command
> >> line
> >>> collect2: error: ld returned 1 exit status
> >>> make[4]: *** [Makefile:25: memembed] Error 1
> >>> make[4]: Leaving directory
> >>>
> >>
> '/apps/amber/18_tools-18_openmpi-3.1.3_gnu-8.2.1/AmberTools/src/packmol_memgen/packmol_memgen/lib/memembed'
> >>> make[3]: *** [Makefile:17: install.memembed] Error 2
> >>> make[3]: Leaving directory
> >>>
> >>
> '/apps/amber/18_tools-18_openmpi-3.1.3_gnu-8.2.1/AmberTools/src/packmol_memgen/packmol_memgen'
> >>
> >> Sigh...the dependence on boost (and perhaps a particular version of
> boost,
> >> not to mention the dependence on python) makes packmol_memgen fragile.
> >> The only suggestion I have is to just comment out the "packmol-memgen"
> >> item in the python_serial target in $AMBERHOME/AmberTools/src/Makefile;
> >> and perhaps try AMBAT, which is an alternative way to build
> >> protein-membrane systems.
> >>
> >> (cc-ing to Stephan Schott in case he recognizes the problem.)
> >>
> >> ...dac
> >>
> >>
>
> --
> Vlad Cojocaru, PD (Habil.), Ph.D.
> -----------------------------------------------
> Project Group Leader
> Department of Cell and Developmental Biology
> Max Planck Institute for Molecular Biomedicine
> Röntgenstrasse 20, 48149 Münster, Germany
> -----------------------------------------------
> Tel: +49-251-70365-324; Fax: +49-251-70365-399
> Email: vlad.cojocaru[at]mpi-muenster.mpg.de
> http://www.mpi-muenster.mpg.de/43241/cojocaru
>
>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
--
Stephan Schott Verdugo
Biochemist
Heinrich-Heine-Universitaet Duesseldorf
Institut fuer Pharm. und Med. Chemie
Universitaetsstr. 1
40225 Duesseldorf
Germany
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Nov 07 2018 - 09:00:02 PST
