I second this - I tried to disable PBSA building for now but of course it is linked into Sander etc so we'd disabling Sander as well.
I'd suggest spending a few hours trying to fix it and if you can't the checkin should be reversed until this is fixed.
Serial builds really aren't a viable option these days - try doing a serial build on a Xeon Phi and you'll understand. :-(
All the best
Ross
> On Feb 18, 2016, at 9:13 AM, Daniel Roe <daniel.r.roe.gmail.com> wrote:
>
> On Wed, Feb 17, 2016 at 6:04 PM, Ray Luo, Ph.D. <ray.luo.uci.edu> wrote:
>> Ross,
>>
>> Please pull again to see whether it fixed the build problem on your
>> side. Looks like multi-thread make does not work ... Too many
>> dependence issues ...
>
> The multi-threaded build definitely doesn't work for me when I hit pbsa stuff:
>
> $ make -j6 install
> ...
> make[2]: Entering directory '/home/droe/Amber/GIT/amber/AmberTools/src/pbsa'
> gcc -c -O3 -mtune=native -fPIC ...
> ...
> Makefile:238: recipe for target 'pb_p3m.LIBPBSA.o' failed
> Makefile:238: recipe for target 'pb_fdfrc.LIBPBSA.o' failed
>
> Amber is pretty large nowadays, and not being able to build in
> parallel is a real drawback. We can currently get around this by
> adding a '.NOTPARALLEL' target to the pbsa Makefile (or forcing -j1),
> but ideally I think the multi-threaded build should be fixed.
>
> -Dan
>
>>
>> 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 Wed, Feb 17, 2016 at 12:59 PM, Ray Luo, Ph.D. <ray.luo.uci.edu> wrote:
>>> Okay, I think I've fixed it. Don't know why the same makefile works in
>>> amber15 but doesn't work in amber16. I'm testing all the
>>> sander/pbsa/nab combs to make sure all can build without interruption
>>> and also pass the tests.
>>>
>>> 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 10:34 PM, Ray Luo, Ph.D. <ray.luo.uci.edu> wrote:
>>>> Finally I can reproduce your problem in building the MPI sander ...
>>>>
>>>> Will get it fixed tomorrow morning ...
>>>>
>>>> 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 10:17 PM, Ray Luo, Ph.D. <ray.luo.uci.edu> wrote:
>>>>> 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
>
>
>
> --
> -------------------------
> Daniel R. Roe, PhD
> Department of Medicinal Chemistry
> University of Utah
> 30 South 2000 East, Room 307
> Salt Lake City, UT 84112-5820
> http://home.chpc.utah.edu/~cheatham/
> (801) 587-9652
> (801) 585-6208 (Fax)
>
> _______________________________________________
> AMBER-Developers mailing list
> AMBER-Developers.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber-developers
_______________________________________________
AMBER-Developers mailing list
AMBER-Developers.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber-developers
Received on Thu Feb 18 2016 - 10:30:03 PST
