RE: Compiling on Linux-PPC970-Myrinet-MPI-xlC

From: Brian Bennion (brian_at_youkai.llnl.gov)
Date: Thu Mar 10 2005 - 18:22:43 CST

I don't know much about xlC especially the header paths.

I did have a similiar if not the same problem and adding
-no_gcc to the compile line got me past my fopen problem.

more likely than not it is a path to the correct system header files. i
don't think it should be looking in g++ though.

Brian

On Fri, 11 Mar 2005, Jordi Camps wrote:

> Hello!
>
> Thank you very much for your fast answer. The define worked, and the
> compilation process advanced until it hit the CollectionMaster.C file. Now,
> when I'm compiling I get the following error:
> ~/MD/NAMD/NAMD_2.5_Source/Linux-PPC-MPI-xlC $ mpicxx -w
> -DNO_STRSTREAM_H -DSOCKLEN_T=size_t
> -I.rootdir/charm-5.8/mpi-linux-gm2-xlc64/include -DCMK_OPTIMIZE=1 -Isrc
> -Iinc -DNAMD_VERSION=\"2.5\" -DNAMD_PLATFORM=\"Linux-PPC-MPI-xlC-MPI\" -O3
> -o obj/CollectionMaster.o -c src/CollectionMaster.C
> "/usr/include/g++/cstdio", line 108.9: 1540-0130 (S) "::fgetpos" is
> not declared.
> "/usr/include/g++/cstdio", line 110.9: 1540-0130 (S) "::fopen" is
> not declared.
> "/usr/include/g++/cstdio", line 115.9: 1540-0130 (S) "::freopen" is
> not declared.
> "/usr/include/g++/cstdio", line 118.9: 1540-0130 (S) "::fsetpos" is
> not declared.
> "/usr/include/g++/cstdio", line 137.9: 1540-0130 (S) "::tmpfile" is
> not declared.
> "src/CollectionMaster.C", line 105.24: 1540-0274 (S) The name lookup
> for "fopen" did not find a declaration.
> "/usr/include/g++/cstdio", line 63.8: 1540-2407 (I) The local macro
> "fopen" is not visible in the current source.
>
> This seems to be a problem with the system libraries, but I think that this
> is quite improbable. Maybe a problem with the include paths :-? Do you have
> some clue?
>
> Thanks!
>
> --
> Jordi Camps Puchades
> Instituto Nacional de Bioinformatica (INB) Nodo Computacional GNHC-2
> UPC-CIRI
> c/. Jordi Girona 1-3
> Modul C6-E201 Tel. : 934 011 650
> E-08034 Barcelona Fax : 934 017 014
> Catalunya (Spain) e-mail: jcamps_at_lsi.upc.edu
>
> -----Mensaje original-----
> De: Gengbin Zheng [mailto:gzheng_at_ks.uiuc.edu]
> Enviado el: jueves, 10 de marzo de 2005 22:55
> Para: Brian Bennion
> CC: Jordi Camps; namd-l_at_ks.uiuc.edu
> Asunto: Re: namd-l: Compiling on Linux-PPC970-Myrinet-MPI-xlC
>
>
>
> Just a small correction:
> -DNO_STRSTREAM_H
>
> The latest NAMD in cvs repository should have this fixed.
>
> Gengbin
>
> Brian Bennion wrote:
>
> >Jordi,
> >
> >Quickly, just block strstream.h from being searched for, its not
> >needed(?)
> >
> >ie in the arch file
> >Linux-PPC-MPI-xlC.arch
> >
> >add
> > -DNO_STREAM_H
> >
> >
> >On Thu, 10 Mar 2005, Jordi Camps wrote:
> >
> >
> >
> >>Hello everybody,
> >>
> >>I'm trying to compile NAMD on a cluster of PPC970 (G5) processors
> >>connected through a Myrinet and using a SuSe Linux 2.6.5 as operating
> >>system. The compiler set that I want to use is the IBM compilers: xlc,
> >>xlC, *_r...
> >>
> >>At this first stage I'm trying to compile only the basic NAMD: no
> >>fftw, no tcl, no plugins.
> >>
> >>I updated the charm++ to the last version (5.8) and it compiled OK and
> >>passed megatest tests with up to 16 processors (8 nodes).
> >>
> >>But when I try to compile the NAMD, I get some errors. The process
> >>followed to compile NAMD has been this:
> >> Update Make.charm
> >> Create files Linux-PPC-MPI-xlC.arch & Linux-PPC-MPI-xlC.base (.fftw,
>
> >>.tcl & .plugins coming soon ;-))
> >> ./config Linux-PPC-MPI-xlC
> >> cd Linux-PPC-MPI-xlC
> >> make
> >>
> >>But the first lines say this:
> >> mkdir inc
> >> mkdir obj
> >> mpicxx -w -DSOCKLEN_T=size_t
> >>-I.rootdir/charm-5.8/mpi-linux-gm2-xlc64/include -DCMK_OPTIMIZE=1
> >>-Isrc -Iinc -DNAMD_VERSION=\"2.5\"
> >>-DNAMD_PLATFORM=\"Linux-PPC-MPI-xlC-MPI\" -O3 -o obj/common.o -c
> src/common.C
> >> "src/InfoStream.h", line 17.10: 1540-0836 (S) The #include file
> >><strstream.h> is not found.
> >> make: *** [obj/common.o] Error 1
> >>
> >>If I edit the InfoStream.h file and change <strstream.h> to
> >><strstream> I get another bunch of messages like:
> >> "src/common.C", line 218.3: 1540-0256 (I) A parameter of type
> >>"std::basic_ostream<char,std::char_traits<char> > &" cannot be
> >>initialized with an expression of type "infostream".
> >> "src/InfoStream.h", line 67.13: 1540-1283 (I) "operator<<(infostream
>
> >>&, const Vector &)" is not a viable candidate."
> >>
> >>Someone has compiled NAMD with xlC? Someone can help me?
> >>
> >>Thanks in advance.
> >>
> >>--
> >>
> >>Jordi Camps Puchades
> >>
> >>Instituto Nacional de Bioinformatica (INB) Nodo Computacional GNHC-2
> >>UPC-CIRI c/. Jordi Girona 1-3
> >>Modul C6-E201 Tel. : 934 011 650
> >>E-08034 Barcelona Fax : 934 017 014
> >>Catalunya (Spain) e-mail: jcamps_at_lsi.upc.edu
> >>
> >>
> >>
> >
> >************************************************
> > Brian Bennion, Ph.D.
> > Bioscience Directorate
> > Lawrence Livermore National Laboratory
> > P.O. Box 808, L-448 bennion1_at_llnl.gov
> > 7000 East Avenue phone: (925) 422-5722
> > Livermore, CA 94550 fax: (925) 424-6605
> >************************************************
> >
> >
>

************************************************
  Brian Bennion, Ph.D.
  Bioscience Directorate
  Lawrence Livermore National Laboratory
  P.O. Box 808, L-448 bennion1_at_llnl.gov
  7000 East Avenue phone: (925) 422-5722
  Livermore, CA 94550 fax: (925) 424-6605
************************************************

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 05:18:38 CST