AMBER Archive (2008)

Subject: Re: AMBER: compiling PMEMD spits out lots of -tp not supported and hangs @ pmemd.f90

From: David Watson (dewatson_at_olemiss.edu)
Date: Fri Dec 05 2008 - 16:30:32 CST

On Dec 5, 2008, at 4:11 PM, Starr Hazard wrote:

> Well I did get to compiling pmemd after successfully compiling amber
> serial and then amber parallel. So in that sense 11.0.074 works for
> me too
>
> When I try "make serial" in the $AMBERHOME/src/pmemd
>
> I see a complaint that it has no rule for target serial

I should have clarified that I am using Amber9.

The "-xHost" option chooses the proper switches for the processor you
are compiling on at the moment (in the ifort 11 suite).

>
>
> Thanks for the suggestion.
>
> Its good to know that 11.0.074 works in your system. At least its
> not a complete Amber-lemon yet :-)
>
> Starr
>
> David Watson wrote:
>> We are running Xeon L5420 dual quads fine with ifort 11.0.074.
>>
>> You might try backing up your config.h and performing a "make
>> clean", regenerate the config.h, and then try to "make serial".
>>
>> On Dec 5, 2008, at 3:15 PM, Robert Duke wrote:
>>
>>
>>> I would be looking at getting either an updated or downlevel (10,
>>> known good release) of the compiler...
>>> Regards - Bob Duke
>>> ----- Original Message ----- From: "Starr Hazard" <hazards_at_musc.edu>
>>> To: <amber_at_scripps.edu>
>>> Sent: Friday, December 05, 2008 4:10 PM
>>> Subject: Re: AMBER: compiling PMEMD spits out lots of -tp not
>>> supported and hangs @ pmemd.f90
>>>
>>>
>>>
>>>> I tried replacing -tpp7 with -xHost in all three cases. It
>>>> whined about overriding xHost with xP and stopped at the same
>>>> place.
>>>>
>>>> Then I changed in the third instance
>>>>
>>>> F90_OPT_HI = -tpp7 -xP -ip -O3 to F90_OPT_HI = -xP -ip -O3
>>>>
>>>> And it stopped at the same place.
>>>>
>>>> Starr
>>>>
>>>> David Watson wrote:
>>>>
>>>>> You might want to change the -tpp7 to -xHost if you are using
>>>>> the ifort 11.0 compiler, although you will only be able to
>>>>> take advantage of the optimizations of the hardware on the
>>>>> machine you are compiling it on.
>>>>> This is fine if you are compiling/running on the same
>>>>> architecture.
>>>>>
>>>>> On Dec 5, 2008, at 2:12 PM, Hazard, E. Starr wrote:
>>>>>
>>>>>
>>>>>
>>>>>> Here's how I set up pmemd
>>>>>>
>>>>>> ./configure linux_em64t ifort mpich2 pubfft nobintraj
>>>>>>
>>>>>>
>>>>>> Here's my config.h
>>>>>>
>>>>>> MATH_DEFINES =
>>>>>> MATH_LIBS =
>>>>>> IFORT_RPATH = /share/apps/intel/lib/intel64:/share/apps/intel/
>>>>>> lib/ intel64:/share/apps/intel/mkl/lib/em64t:/share/apps/
>>>>>> mpich2- intel_11_install/lib:/share/apps/ccp4-6.0.2/lib:/opt/
>>>>>> openmpi/ 1.1.4/ lib:/opt/lava/6.1/linux2.6-glibc2.3-ia32e/lib
>>>>>> MATH_DEFINES = -DMKL
>>>>>> MATH_LIBS = /share/apps/intel/mkl/lib/em64t/
>>>>>> libmkl_intel_lp64.a / share/apps/intel/mkl/lib/em64t/
>>>>>> libmkl_core.a /share/apps/intel/mkl/ lib/em64t/
>>>>>> libmkl_sequential.a
>>>>>> FFT_DEFINES = -DPUBFFT
>>>>>> FFT_INCLUDE =
>>>>>> FFT_LIBS =
>>>>>> NETCDF_HOME = /home/hazards/COMPILE/amber10/src/netcdf
>>>>>> NETCDF_DEFINES = -DBINTRAJ
>>>>>> NETCDF_MOD = netcdf.mod
>>>>>> NETCDF_LIBS = $(NETCDF_HOME)/lib/libnetcdf.a
>>>>>> MPI_HOME = /share/apps/mpich2-intel_11_install
>>>>>> MPI_DEFINES = -DMPI -DSLOW_NONBLOCKING_MPI
>>>>>> MPI_INCLUDE = -I$(MPI_HOME)/include
>>>>>> MPI_LIBDIR = $(MPI_HOME)/lib
>>>>>> MPI_LIBS = -L$(MPI_LIBDIR) -lmpich -lrt -luuid -lpthread
>>>>>> DIRFRC_DEFINES = -DDIRFRC_EFS -DDIRFRC_COMTRANS -DDIRFRC_NOVEC
>>>>>> CPP = /lib/cpp
>>>>>> CPPFLAGS = -traditional -P
>>>>>> F90_DEFINES = -DFFTLOADBAL_2PROC
>>>>>>
>>>>>> F90 = ifort
>>>>>> MODULE_SUFFIX = mod
>>>>>> F90FLAGS = -c -auto
>>>>>> F90_OPT_DBG = -g -traceback
>>>>>> F90_OPT_LO = -tpp7 -O0
>>>>>> F90_OPT_MED = -tpp7 -O2
>>>>>> F90_OPT_HI = -tpp7 -xP -ip -O3
>>>>>> F90_OPT_DFLT = $(F90_OPT_HI)
>>>>>>
>>>>>> CC = gcc
>>>>>> CFLAGS =
>>>>>>
>>>>>> LOAD = ifort
>>>>>> LOADFLAGS =
>>>>>> LOADLIBS = -limf -lsvml -Wl,-rpath=$(IFORT_RPATH)
>>>>>>
>>>>>>
>>>>>> make install hums along nicely up to this point and then hangs
>>>>>> here for a LONG time (up to 20 minutes)
>>>>>>
>>>>>> /lib/cpp -traditional -P -I/share/apps/mpich2-
>>>>>> intel_11_install/ include -DPUBFFT -DBINTRAJ -DMPI -
>>>>>> DSLOW_NONBLOCKING_MPI - DDIRFRC_EFS -DDIRFRC_COMTRANS -
>>>>>> DDIRFRC_NOVEC -DMKL - DFFTLOADBAL_2PROC pmemd.fpp pmemd.f90
>>>>>> ifort -c -auto -tpp7 -xP -ip -O3 pmemd.f90
>>>>>> ifort: command line remark #10148: option '-tp' not supported
>>>>>>
>>>>>>
>>>>>> should I just keep waiting?
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Then hangs.... not for 5 minutes but up to 20.
>>>>>>
>>>>>>
>>>>> 

--
David Watson • Graduate Student • Medicinal Chemistry • 425 Faser Hall  
• University MS  38677 • 1.662.915.1663

-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" (in the *body* of the email)
      to majordomo_at_scripps.edu