Hi Jason, thanks for your time.
to check which patches ardeady been applied i used
./patch_amber.py --patch-level command.
then i confirm the updates by
./patch_amber.py --check-updates [--verbose | --quiet]
./patch_amber.py --update-tree to download and apply.
in the end everything looks fine.at the end when i apply
cd $AMBERHOME
./configure -mpi gnu
everything is fine till the configuration but while doing installation
make install
it gives following error.
Cleaning the src directories. This may take a few moments.
Configure complete.
[nuria.zinc ~/amber12]$ make install
cd AmberTools/src && make install
make[1]: Entering directory `/home/nuria/amber12/AmberTools/src'
Starting installation of AmberTools12 MPI at Tue Jan 15 15:22:44 KST 2013.
(cd etc && make parallel)
make[2]: Entering directory `/home/nuria/amber12/AmberTools/src/etc'
mpif90 -DBINTRAJ -DMPI -o numprocs numprocs.F90
make[2]: mpif90: Command not found
make[2]: *** [parallel] Error 127
make[2]: Leaving directory `/home/nuria/amber12/AmberTools/src/etc'
make[1]: *** [parallel] Error 2
make[1]: Leaving directory `/home/nuria/amber12/AmberTools/src'
make: *** [install] Error 2
Nuria
PhD student
________________________________
From: Jason Swails <jason.swails.gmail.com>
To: Bio Learner <bio.learner.yahoo.com>; AMBER Mailing List <amber.ambermd.org>
Sent: Tuesday, January 15, 2013 12:56 AM
Subject: Re: [AMBER] installation error
On Mon, Jan 14, 2013 at 4:58 AM, Bio Learner <bio.learner.yahoo.com> wrote:
>
>
> hello, im trying to make my installation in parallel version...for that i
> come across following error. can anyone tell me what is this and how to
> cope it?
>
-cuda is not the same as what we call "parallel". While CUDA-enabled GPUs
are technically highly parallel, the 'parallel' code in Amber is based on
the message passing interface (MPI). Therefore, to build in 'parallel',
you should use
./configure -mpi intel
rather than
./configure -cuda intel
That said, I highly suggest that you become familiar and comfortable with
the serial version of Amber, and with Unix commands in general, before
moving on to more advanced options like GPU and parallel processing.
>
> [nuria.zinc ~/amber12]$ ./configure -cuda intel
> Checking for updates...
> Amber12 is up to date
> AmberTools12 has 3 unapplied patches: 31, 32, 33
>
> There are patches available. Do you want to apply them now? [y/n]
> (Recommended Y)
> y
> Determining patches to download for AmberTools12
> Downloading bugfix.31
> Downloading bugfix.32
> Downloading bugfix.33
>
> Determining patches to download for Amber12
> Amber12 is up to date. Nothing to download
>
> Applying AmberTools12 patches
> Error or warning during patching process for
> .patches/AmberTools_Unapplied_Patches/bugfix.31:
> Reversed (or previously applied) patch detected! Skipping patch.
> 1 out of 1 hunk ignored -- saving rejects to file
> AmberTools/src/cpptraj/src/Action_AutoImage.cpp.rej
> Reversed (or previously applied) patch detected! Skipping patch.
> 1 out of 1 hunk ignored -- saving rejects to file
> AmberTools/src/cpptraj/src/Action_Hbond.cpp.rej
> 1 out of 1 hunk FAILED -- saving rejects to file
> AmberTools/src/cpptraj/src/Action_Image.cpp.rej
> Reversed (or previously applied) patch detected! Skipping patch.
> 1 out of 1 hunk ignored -- saving rejects to file
> AmberTools/src/cpptraj/src/ArgList.cpp.rej
> Reversed (or previously applied) patch detected! Skipping patch.
> 1 out of 1 hunk ignored -- saving rejects to file
> AmberTools/src/cpptraj/src/ArgList.h.rej
> Reversed (or previously applied) patch detected! Skipping patch.
> 1 out of 1 hunk ignored -- saving rejects to file
> AmberTools/src/cpptraj/src/CpptrajState.cpp.rej
> Reversed (or previously applied) patch detected! Skipping patch.
> 1 out of 1 hunk ignored -- saving rejects to file
> AmberTools/src/cpptraj/src/CpptrajState.h.rej
> Reversed (or previously applied) patch detected! Skipping patch.
> 1 out of 1 hunk ignored -- saving rejects to file
> AmberTools/src/cpptraj/src/DistRoutines.cpp.rej
> 1 out of 2 hunks FAILED -- saving rejects to file
> AmberTools/src/cpptraj/src/Frame.cpp.rej
> 1 out of 1 hunk FAILED -- saving rejects to file
> AmberTools/src/cpptraj/src/Frame.h.rej
> Reversed (or previously applied) patch detected! Skipping patch.
> 3 out of 3 hunks ignored -- saving rejects to file
> AmberTools/src/cpptraj/src/Traj_CharmmDcd.cpp.rej
> 4 out of 5 hunks FAILED -- saving rejects to file
> AmberTools/src/cpptraj/src/main.cpp.rej
>
It seems like the patches were previously applied, but for some reason they
were not logged. You can fix this by running the command:
cd $AMBERHOME && ./patch_amber.py --update --ignore-fail
> Automatic patching failed! Check the errors before re-configuring
> NOT updating your tree and continuing anyway.
> Searching for python2... Found python2.4: /usr/bin/python2.4
> Error: CUDA_HOME is not set. This must point to your NVIDIA tools
> installation
> Configure failed due to the errors above!
>
As Dan said, the messages above tell you what you need to do in order to
fix the CUDA build, if that is actually what you want.
Good luck,
Jason
--
Jason M. Swails
Quantum Theory Project,
University of Florida
Ph.D. Candidate
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
Received on Mon Jan 14 2013 - 22:30:02 PST
