On Fri, Nov 12, 2010 at 2:16 PM, Dean Cuebas
<deancuebas.missouristate.edu>wrote:
> Dear users,
>
> Thanks very much for your help regarding compiling on Snow Leopard!
>
> Just wanted to clarify that
>
> > sudo gcc_select gcc-mp-4.4
>
> Didn't work, as the only compilers available using
>
> sudo gcc_select -l
>
> did not give gcc-mp.4.4, but did give
>
> mp-gcc44
>
You are right, this was my mistake. My memory is not so precise, I suppose.
>
> So I chose that (and I saw in a previous email on the list, that you
> (Jason)
> selected that also...)
>
Yes, this is the one I use (I'm checking out 4.5 now, since it's in its
supposedly "stable" stage, but there are no performance differences in the
limited testing I've done with pmemd).
>
> After ./configure -macAccelerate gnu and make install I get the following:
> ++++++++++++
> .
> .
> .
>
> ranlib arpack.a
> mv arpack.a /usr/local/amber11/lib
> gcc -c -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE -DBINTRAJ -o main.o
> main.c
> gcc -c -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE -DBINTRAJ -o rdparm.o
> rdparm.c
> gcc -c -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE -DBINTRAJ -o
> dispatch.o
> dispatch.c
> gcc -c -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE -DBINTRAJ -o help.o
> help.c
> gcc -c -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE -DBINTRAJ -o utility.o
> utility.c
> gcc -c -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE -DBINTRAJ -o second.o
> second.c
> gcc -c -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE -DBINTRAJ -o io.o io.c
> io.c:782: error: conflicting types for 'gzipFileSize'
> io.c:681: note: previous definition of 'gzipFileSize' was here
> io.c:814: error: conflicting types for 'bzip2FileSize'
> io.c:728: note: previous definition of 'bzip2FileSize' was here
> io.c:840: error: conflicting types for 'zipFileSize'
> io.c:755: note: previous definition of 'zipFileSize' was here
> make[1]: *** [io.o] Error 1
> make: *** [serial] Error 2
> ++++++++++++++++++++++++++++++++++++++
>
> Any reason for the errors???
>
Maybe there is leftover *stuff* from a previous installation attempt? Did
you make clean? Is this a fresh checkout with the bug fixes applied? My
best guess would probably be that something *bad* happened applying bug
fixes (specifically 4 and 9). bugfix.4 makes initial changes to io.c
involving gzip/bzip sizes, and bugfix.9 adjusts those changes, too, so if
care is not taken it's probably pretty easy to mangle things, quite easily
giving rise to these errors that you're seeing. This could happen, for
instance, if you initially applied bug fixes, and then applied some again,
since I don't know how smart patch would be when it tries to pass over the
same chunk of code twice again after it's already passed over it twice to
start with.
My advice, rm -fr your whole amber11 directory (or maybe just
amber/AmberTools), unpack a fresh copy, and apply the bug fixes before doing
anything else.
Hope this helps,
Jason
> Make test shows a bunch of errors involving arpack
>
> Any suggestions?
>
> Thanks so much for your help!
>
> Dean
>
> PS. The above is compiling after applying the bugfix.all to the source
>
>
> Dr. Dean Cuebas, Associate Professor of Chemistry
> deancuebas.missouristate.edu, Ph 417-836-8567 FAX 417-836-5507
> Dept. of Chemistry, Missouri State University
> Springfield, Missouri 65897
>
>
>
> > From: Jason Swails <jason.swails.gmail.com>
> > Reply-To: AMBER Mailing List <amber.ambermd.org>
> > Date: Thu, 11 Nov 2010 10:36:37 -0600
> > To: AMBER Mailing List <amber.ambermd.org>
> > Subject: Re: [AMBER] Compiling Ambertools 1.4 on MacBook
> SnowLeopard10.6.4
> >
> > Hi Dean,
> >
> > I'm guessing this is the "hpc.sourceforge" site, and for some reason it
> > looks like they like to have only the bleeding-edge version of any GCC
> > compiler on there. Last time I tried downloading from there, it was 4.5,
> > experimental, now 4.5 is official release and 4.6 is experimental. I
> would
> > avoid using these compilers, as they're most likely to be buggy over
> > official releases.
> >
> > My far more preferred method is using MacPorts, which in my experience
> has
> > caught up to advances in Snow Leopard much more than alternatives (i.e.
> > Fink). Download MacPorts and install a GCC from there. I currently have
> > 4.3 through 4.6 installed, but 4.4 seems to work the best for me.
> >
> > sudo port -v install gcc44
> > sudo port -v install gcc_select
> > sudo gcc_select gcc-mp-4.4
> >
> > The above commands should install GCC 4.4.5, install a selection utility
> > that makes selecting a version of GCC compilers very easy, then using
> that
> > utility to select your newly installed GCC. These are native x86_64
> > binaries, unlike Fink's versions that I've seen (which are i386, but can
> > build x86_64 objects if explicitly passed -m64).
> >
> > Then you should be ready to build without hiccups.
> >
> > Hope this helps,
> > Jason
> >
> > On Thu, Nov 11, 2010 at 10:19 AM, Dean Cuebas
> > <deancuebas.missouristate.edu>wrote:
> >
> >> Dear users,
> >>
> >> I've tried to install Ambertools1.4 and have a compiler problem.
> >> I followed the instructions on amberonmac.blogspot.com SATURDAY, MAY
> 15,
> >> 2010 post.
> >>
> >>
> >> 1) Dowloaded and installed gcc-snwleo-intel-bin.tar.gz
> >> 2) type gcc -v and I get:
> >> ++++++++++++++++++++++++
> >> Using built-in specs.
> >> COLLECT_GCC=gcc
> >>
> >>
> COLLECT_LTO_WRAPPER=/usr/local/libexec/gcc/x86_64-apple-darwin10/4.6.0/lto-w
> >> rapper
> >> Target: x86_64-apple-darwin10
> >> Configured with: ../gcc-4.6-20100703/configure
> >> --enable-languages=fortran,c++ --host=x86_64-apple-darwin10
> >> --build=x86_64-apple-darwin10
> >> Thread model: posix
> >> gcc version 4.6.0 20100703 (experimental) (GCC)
> >> ++++++++++++++++++++++++++++++++++
> >> Which is a higher version of gcc than in the amberonmac demo.
> >>
> >> 3) Type gfortran -v and I get:
> >>
> >> +++++++++++++++++++++++++++++
> >> Using built-in specs.
> >> COLLECT_GCC=gfortran
> >>
> >>
> COLLECT_LTO_WRAPPER=/usr/local/libexec/gcc/x86_64-apple-darwin10/4.6.0/lto-w
> >> rapper
> >> Target: x86_64-apple-darwin10
> >> Configured with: ../gcc-4.6-20100703/configure
> >> --enable-languages=fortran,c++ --host=x86_64-apple-darwin10
> >> --build=x86_64-apple-darwin10
> >> Thread model: posix
> >> gcc version 4.6.0 20100703 (experimental) (GCC)
> >> ++++++++++++++++++++++++++++++++++++
> >>
> >> 4) I cd to the src directory of Ambertools and type
> >> ./configure -macAccelerate gnu
> >>
> >> +++++++++++++++++++++++++++++++++++++++++++++
> >> Your AMBERHOME environment variable should be set to /usr/local/amber11
> >>
> >>
> >> Obtaining the gnu suite version:
> >> gcc -v
> >> The version is 4.6.0
> >>
> >> Testing the gcc compiler:
> >> gcc -DUSE_AMBER_C9XCOMPLEX -D_FILE_OFFSET_BITS=64
> -D_LARGEFILE_SOURCE
> >> -o testp testp.c
> >> testp.c:1:19: fatal error: stdio.h: No such file or directory
> >> compilation terminated.
> >> ./configure: line 967: ./testp: No such file or directory
> >> Error: Unable to compile a C program using gcc -DUSE_AMBER_C9XCOMPLEX
> >> -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
> >> Please check your compiler settings or configure flags.
> >> ++++++++++++++++++++++++++++++++++++++++++++++
> >>
> >> Any help would be appreciated!
> >>
> >> Thanks in advance.
> >>
> >> Dean
> >> --
> >> Dr. Dean Cuebas, Associate Prof of Chemistry
> >> deancuebas.missouristate.edu, Ph 417-836-8567 FAX 417-836-5507
> >> Dept. of Chemistry, Missouri State University
> >> Springfield, Missouri 65897
> >>
> >>
> >>
> >> _______________________________________________
> >> AMBER mailing list
> >> AMBER.ambermd.org
> >> http://lists.ambermd.org/mailman/listinfo/amber
> >>
> >
> >
> >
> > --
> > Jason M. Swails
> > Quantum Theory Project,
> > University of Florida
> > Ph.D. Graduate Student
> > 352-392-4032
> > _______________________________________________
> > AMBER mailing list
> > AMBER.ambermd.org
> > http://lists.ambermd.org/mailman/listinfo/amber
>
>
>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
--
Jason M. Swails
Quantum Theory Project,
University of Florida
Ph.D. Graduate Student
352-392-4032
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Fri Nov 12 2010 - 13:00:03 PST
