Hi Adelene,
I'm not sure which instructions you used from the "AMBER on Mac" site,
but I've been able to compile AmberTools 1.4 (and Amber 11) on two
Snow Leopard machines without any issues using a very slight variation
on these instructions: http://amberonmac.blogspot.com/2010/04/step-by-step-ambertools-14-installation.html
. Here is an abbreviated version of what I do:
(1) download and install the latest Xcode (3.2.x)
(2) download and install the latest version of MacPorts (1.9.x)
(3) use MacPorts to install gcc 4.4.4 (sudo port install gcc44)
(4) use MacPorts to install gcc_select (sudo port install gcc_select)
(5) use gcc_select to make gcc 4.4.4 your default compilers (sudo
gcc_select mp-gcc44)
(6) download AmberTools and unpack it into a directory *other than* /
usr/local (e.g., /Users/your_username/amber11) -- this avoids all of
the permission/ownership issues of trying to build in /usr/local
(7) set your $AMBERHOME environment variable to the appropriate
directory (this depends on what shell you're using, but usually can be
done in the .profile file in your home directory)
(8) configure AmberTools as "usual" (./configure -macAccelerate gnu)
In my opinion, setting up AmberTools this way avoids a lot of the
pitfalls of the pre-built compilers. While they can work well, it is
also possible that your machine is different enough from the one the
compilers were built on that you can run into trouble.
Best,
Paul
On Jul 26, 2010, at 9:37 AM, Adelene Sim wrote:
> Thank you for the patch. I've tried installing Amber Tools 11 on
> snow leopard, based on the directions here http://amberonmac.blogspot.com/
> (both with and without patch used), and get this error:
>
> Undefined symbols:
> "_rfftwnd_f77_destroy_plan_", referenced from:
> ___rism3d_c_MOD_reallocbox in libnab.a(rism3d_c.o)
> ___rism3d_c_MOD_reallocbox in libnab.a(rism3d_c.o)
> ___rism3d_c_MOD_rism3d_destroy in libnab.a(rism3d_c.o)
> ___rism3d_c_MOD_rism3d_destroy in libnab.a(rism3d_c.o)
> "_rfftwnd_f77_one_real_to_complex_", referenced from:
> ___rism_fft_MOD_rlft3i in libnab.a(rism_fft.o)
> "_rfftw3d_f77_create_plan_", referenced from:
> ___rism_fft_MOD_fftw_make_plans in libnab.a(rism_fft.o)
> ___rism_fft_MOD_fftw_make_plans in libnab.a(rism_fft.o)
> "_rfftwnd_f77_one_complex_to_real_", referenced from:
> ___rism_fft_MOD_rlft3i in libnab.a(rism_fft.o)
> ld: symbol(s) not found
> collect2: ld returned 1 exit status
> cc failed!
> make[1]: *** [matgen] Error 1
> make: *** [serial] Error 2
>
>
> Do you know what could be wrong?
>
> Thanks.
>
> Adelene
>
> ----- Original Message -----
> From: "case" <case.biomaps.rutgers.edu>
> To: "AMBER Mailing List" <amber.ambermd.org>
> Sent: Monday, July 26, 2010 7:06:27 AM
> Subject: Re: [AMBER] compile ambertools on mac
>
> On Sun, Jul 25, 2010, Mengjuei Hsieh wrote:
>>
>> I confirm that fftw2 could not be properly linked at least in Mac
>> OS X
>> 10.4.x. This is because the configure script of fftw2 does not handle
>> later versions of gfortran built on 10.4.x and subsequently bails on
>> testing with disabled fortran objects. A detail analysis shows that
>> in
>> my 10.4.x box, gfortran lists -lcrt1.o as one of the system library
>> but the configure script doesn't like it.
>>
>> So a temporary solution would be to apply my patch (attached) in
>> $AMBERHOME, and re-do the configure.
>> $ cd $AMBERHOME; patch -p0 -N < fftw_prob.patch
>> $ cd AmberTools/src; ./configure -macAccelerate gnu
>
> Thanks. Two quick questions: is there a reason this patch should
> not be
> applied in general? Second, is there any reason not to make the
> same change
> at line 1315 as you make at line 1319?
>
> ....dac
>
>
> _______________________________________________
> 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
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Mon Jul 26 2010 - 10:00:07 PDT
