David,
I did what you suggested and recomplied sander.MPI but the test jobs
cannot be passed with error messages "MKL ERROR: Parameter 1 was
incorrect on entry to vdExp"
The library line in config.h is shown below:
LOADLIB= -L/opt/openmpi-1.4.2-intel64/lib -lmpi_f90 -lmpi_f77 -lmpi
-lopen-rte -lopen-pal -ldl -Wl,--export-dynamic -lnsl -lutil
-L/opt/intel/Compiler/11.1/072/mkl/lib/em64t -lmkl_intel_ilp64
-lmkl_intel_thread -lmkl_core -lsvml -openmp -lguide -lpthread
Did I make some mistakes there?
Thanks a lot.
Yubo
On Thu, 22 Jul 2010 11:22:56 -0500
David Watson <dewatson.olemiss.edu> wrote:
> You will need to edit the config.h file after each ./configure so
>that it has the appropriate link line if you have newer Intel
>compilers.
>
>For an exact link line, please refer to:
> http://software.intel.com/en-us/articles/intel-mkl-link-line-advisor/
>
> Also, please make sure that you source the proper
>ifortvars.sh/iccvars.sh/mklvars*.sh before you attempt to "make"
>Amber.
> It appears you have done what is necessary for the serial version,
>so I think the link line advisor is the place to visit first.
>
> On Jul 22, 2010, at 10:43 AM, Yubo Fan wrote:
>
>> Jason,
>>
>> The compiling still failed with error message "ld: cannot find
>>-lmkl".
>> I've already changed lvml to lsvml as you instructed. I think
>>MPI_HOME
>> were correctly set. Is it possible that the Intel Compiler I'm using
>> is too new? Any advice?
>>
>> On Tue, 29 Jun 2010 15:27:29 -0400
>> Jason Swails <jason.swails.gmail.com> wrote:
>>> Hello,
>>>
>>> Amber9 attempted to manually set the mpi libraries rather than
>>> trusting the
>>> MPI wrappers. Assuming you have properly installed openMPI, you can
>>> do the
>>> following: replace all ifort with mpif90 in the config.h file. You
>>> can also
>>> get rid of all of the MPI_LIB or MPI_LIBRARY definitions, since
>>> mpif90 takes
>>> care of that implicitly. Also make sure MPI_HOME is set (although
>>> this MAY
>>> not be necessary). MPI_HOME I think is only required if the
>>> preprocessor
>>> needs to know about mpif.h, in which case mpif90 doesn't take care
>>> of that.
>>> MPI_HOME should be set to the directory containing bin/mpif90. For
>>> example,
>>> if "which mpif90" returns /usr/local/bin/mpif90, then MPI_HOME
>>> should be
>>> /usr/local.
>>>
>>> Good luck!
>>> Jason
>>>
>>> On Tue, Jun 29, 2010 at 2:41 PM, Yubo Fan
>>> <yubofan.mail.chem.tamu.edu>wrote:
>>>
>>>> Jason,
>>>>
>>>> The serial compilation worked now. But the parallel one did not with
>>>> error message:
>>>>
>>>>> ./configure -openmpi -bintraj ifort_x86_64
>>>>> make clean
>>>>> make parallel
>>>> ================
>>>> make[3]: Leaving directory `/opt/amber9/src/netcdf/src'
>>>> Not making `cxx/install' because no C++ compiler
>>>> make[2]: Leaving directory `/opt/amber9/src/netcdf/src'
>>>> cp ../netcdf/include/*.mod .
>>>> cpp -traditional -Iopenmpi-1.4.2-intel64/include -P -DMPI -DMKL
>>>> -DBINTRAJ evb_vars.f > _evb_vars.f
>>>> ifort -c -w95 -mp1 -O0 -FR -o evb_vars.o _evb_vars.f
>>>> cpp -traditional -Iopenmpi-1.4.2-intel64/include -P -DMPI -DMKL
>>>> -DBINTRAJ evb_input.f > _evb_input.f
>>>> ifort -c -w95 -mp1 -O0 -FR -o evb_input.o _evb_input.f
>>>> cpp -traditional -Iopenmpi-1.4.2-intel64/include -P -DMPI -DMKL
>>>> -DBINTRAJ evb_init.f > _evb_init.f
>>>> evb_init.f:18: error: mpif.h: No such file or directory
>>>> make[1]: *** [evb_init.o] Error 1
>>>> make[1]: Leaving directory `/opt/amber9/src/sander'
>>>> make: *** [parallel] Error 2
>>>> ================
>>>>
>>>> BTW, the environment setting is good for the compilation of Amber11.
>>>> Any advice?
>>>>
>>>> Best,
>>>> Yubo
>>>>
>>>>
>>>> On Tue, 29 Jun 2010 09:46:24 -0400
>>>> Jason Swails <jason.swails.gmail.com> wrote:
>>>>> Edit the config.h file and change -lvml to -lsvml wherever it
>>>> occurs
>>>>> in that
>>>>> file. Then try again.
>>>>>
>>>>> Good luck!
>>>>> Jason
>>>>>
>>>>> On Tue, Jun 29, 2010 at 9:39 AM, Yubo Fan
>>>>> <yubofan.mail.chem.tamu.edu>wrote:
>>>>>
>>>>>> Hi, everyone,
>>>>>>
>>>>>> I tried to compile Amber 9 and it failed with the following
>>>> message:
>>>>>>
>>>>>>> ./configure -bintraj ifort_x86_64
>>>>>>> make clean
>>>>>>> make serial
>>>>>>
>>>>>> =====
>>>>>> make[2]: `nxtsec.o' is up to date.
>>>>>> make[2]: Leaving directory `/opt/amber9/src/lib'
>>>>>> ifort -o addles lesmain.o addspace.o readprm.o writprm.o
>>>>>> readcrd.o
>>>>>> writcrd.o pick.o rline.o nxt.o intgr.o find.o
>>>> of.o
>>>>>> geti.o unit.o getc.o alert.o echo.o get4c.o getd.o
>>>>>> wlesprm.o lesprm.o les2prm.o checksz.o ../lib/random.o
>>>>>> ../lib/mexit.o ../lib/nxtsec.o
>>>>>> -L/opt/intel/Compiler/11.1/072/mkl/lib/em64t -lvml -lmkl_lapack
>>>>>> -lmkl
>>>>>> -lguide -lpthread
>>>>>> ld: cannot find -lvml
>>>>>> make[1]: *** [addles] Error 1
>>>>>> make[1]: Leaving directory `/opt/amber9/src/addles'
>>>>>> make: *** [serial] Error 2
>>>>>> =====
>>>>>>
>>>>>> The compiling of Amber 11 was succeeded. The intel compiler is
>>>>>> version
>>>>>> 11. Is it too new? Any advice?
>>>>>>
>>>>>> Sincerely,
>>>>>>
>>>>>> Yubo Fan
>>>>>> Computational Chemistry and Drug Design Lab
>>>>>> Bioengineering and Bioinformatics Program
>>>>>> The Methodist Hospital Research Institute (TMHRI)
>>>>>> 6565 Fannin St B5-022
>>>>>> Houston, TX 77030
>>>>>> Email: yfan.tmhs.org
>>>>>> Phone: (713) 441-8681
>>>>>>
>>>>>> _______________________________________________
>>>>>> 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
>>>>
>>>> Yubo Fan
>>>> Computational Chemistry and Drug Design Lab
>>>> Bioengineering and Bioinformatics Program
>>>> The Methodist Hospital Research Institute (TMHRI)
>>>> 6565 Fannin St B5-022
>>>> Houston, TX 77030
>>>> Email: yfan.tmhs.org
>>>> Phone: (713) 441-8681
>>>>
>>>> _______________________________________________
>>>> 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
>>
>> Sincerely,
>>
>> Yubo Fan
>> Computational Chemistry and Drug Design Lab
>> Bioengineering and Bioinformatics Program
>> The Methodist Hospital Research Institute (TMHRI)
>> 6565 Fannin St B5-022
>> Houston, TX 77030
>> Email: yfan.tmhs.org
>> Phone: (713) 441-8681
>>
>> _______________________________________________
>> 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
Yubo Fan
Computational Chemistry and Drug Design Lab
Bioengineering and Bioinformatics Program
The Methodist Hospital Research Institute (TMHRI)
6565 Fannin St B5-022
Houston, TX 77030
Email: yfan.tmhs.org
Phone: (713) 441-8681
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Fri Jul 23 2010 - 07:30:04 PDT
