Re: [AMBER-Developers] Sander parallel build broken by PBSA updates

From: Ray Luo, Ph.D. <ray.luo.uci.edu>
Date: Tue, 16 Feb 2016 22:17:25 -0800

Hi Ross,

Looks like that there is no extra untracked file at least in all the
src folders, though there were indeed two .swp files in the test
cases. So I think lack of "git clean -fdx" is not the problem. I've
finished the serial build and is running "test.sander.BASIC". Looking
good, so far.

I'll do the MPI build next after another "git clean -fdx" ...

However, I think one possible reason could be the use of multi-thread
in make. There may be inter-dependent issues involved.

I'll do some experiment with the MPI build to see whether this is the cause.

Ray
--
Ray Luo, Ph.D.
Professor
Biochemistry, Molecular Biophysics, Chemical Physics,
Chemical and Biomedical Engineering
University of California, Irvine, CA 92697-3900
On Tue, Feb 16, 2016 at 8:24 PM, Ross Walker <ross.rosswalker.co.uk> wrote:
> Hi Ray,
>
> It's not about whether you compiled in that folder - it is whether you maybe edited files in that folder that you never committed / pushed. Try doing:
>
> git clean -f -d -x
> git status
>
> And see what it reports.
>
> All the best
> Ross
>
>> On Feb 16, 2016, at 23:04, Ray Luo <rluo.uci.edu> wrote:
>>
>> Hi Ross,
>>
>> Yes, I think my folder is clean since I have never compiled anything
>> in my amber git folder. Every time after syncing with the master
>> branch, I copy the whole master branch to a new folder for compiling
>> and testing.
>>
>> I'm about to test this evening's synced version, but will do "git
>> clean -fxd" before configuring/compiling. Maybe this is why.
>>
>> By the way, could you add my uci email to the mailing list as well,
>> i.e. "ray.luo.uci.edu"?
>>
>> All the best,
>> Ray
>> --
>> Ray Luo, Ph.D.
>> Professor
>> Biochemistry, Molecular Biophysics, Chemical Physics,
>> Chemical and Biomedical Engineering
>> University of California, Irvine, CA 92697-3900
>>
>>
>> On Tue, Feb 16, 2016 at 7:51 PM, Ross Walker <ross.rosswalker.co.uk> wrote:
>>> Hi Ray,
>>>
>>> If I do this - on my OSX laptop (GCC 4.9.3), MPICH version 3.1.4
>>>
>>> git pull
>>> git clean -f -d -x
>>> git status
>>> On branch master
>>> Your branch is up-to-date with 'origin/master'.
>>> nothing to commit, working directory clean
>>>
>>> ./configure -mpi gnu
>>> make install
>>>
>>> ...
>>> ...
>>> dnrm2.F90:113.24:
>>>
>>>      ASSIGN 110 TO NEXT
>>>                        1
>>> Warning: Deleted feature: ASSIGN statement at (1)
>>> mpif90  -DBINTRAJ -DEMIL -DMPI    -c -O3 -mtune=native -fPIC  -ffree-form  -I/Users/rcw/amber/amber/include -I/Users/rcw/amber/amber/include    -o dcopy.o dcopy.F90
>>> mpif90  -DBINTRAJ -DEMIL -DMPI    -c -O3 -mtune=native -fPIC  -ffree-form  -I/Users/rcw/amber/amber/include -I/Users/rcw/amber/amber/include    -o pb_force.o pb_force.F90
>>> pb_force.F90:292.77:
>>>
>>>                                   sa_init, sa_driver, sa_free, sa_free_mb,  &
>>>                                                                           1
>>> Error: Symbol 'saslave_init' referenced at (1) not found in module 'solvent_accessibility'
>>> make[3]: *** [pb_force.o] Error 1
>>> make[2]: *** [libpbsa] Error 2
>>> make[1]: *** [parallel] Error 2
>>> make: *** [install] Error 2
>>>
>>> Same problem. Are you sure you have the latest tree and it is clean and doesn't have any modified files locally or files you forgot to add?
>>>
>>> All the best
>>> Ross
>>>
>>>
>>>> On Feb 16, 2016, at 15:56, Ray Luo, Ph.D. <ray.luo.uci.edu> wrote:
>>>>
>>>> Hi Ross,
>>>>
>>>> So far I can "configure -mpi gnu", "make install", and can also pass
>>>> "make sander.BASIC.MPI". I'm running "make test.parallel.MM". I'm
>>>> using "mpirun -np 4" for the mpi jobs. This is on Rocks 6.1/Centos
>>>> 6.1. Which step was the problem?
>>>>
>>>> However I disabled "pbsa_mbfocus" in AmberTools/test/ since I'm in the
>>>> process of removing this feature to accommodate incoming CUDA. I also
>>>> need to update the test cases for sander/nab/mmpbsa ... which will be
>>>> partially in today for sander/nab. I'm trying to get mmpbsa.py to work
>>>> with the new python environment.
>>>>
>>>> All the best,
>>>> Ray
>>>> --
>>>> Ray Luo, Ph.D.
>>>> Professor
>>>> Biochemistry, Molecular Biophysics, Chemical Physics,
>>>> Chemical and Biomedical Engineering
>>>> University of California, Irvine, CA 92697-3900
>>>>
>>>>
>>>> On Tue, Feb 16, 2016 at 10:21 AM, Ray Luo, Ph.D. <ray.luo.uci.edu> wrote:
>>>>> Hi Ross,
>>>>>
>>>>> I'm looking into this ... There was a major overhaul of the code in
>>>>> addition to new features in the last check in.
>>>>>
>>>>> All the best,
>>>>>
>>>>> Ray
>>>>>
>>>>> --
>>>>> Ray Luo, Ph.D.
>>>>> Professor
>>>>> Biochemistry, Molecular Biophysics, Chemical Physics,
>>>>> Chemical and Biomedical Engineering
>>>>> University of California, Irvine, CA 92697-3900
>>>
>
_______________________________________________
AMBER-Developers mailing list
AMBER-Developers.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber-developers
Received on Tue Feb 16 2016 - 22:30:04 PST
Custom Search