Custom Search
-- Ray Luo, Ph.D. Professor Biochemistry, Molecular Biophysics, Chemical Physics, Chemical and Biomedical Engineering University of California, Irvine, CA 92697-3900 On Thu, Feb 18, 2016 at 11:08 AM, Gerald Monard <Gerald.Monard.univ-lorraine.fr> wrote: > Hi, > > I think that I found a fix. The problem was in the makedepend program in > pbsa/ that had not been updated. That should work now. Can anybody test? > > Best, > > Gerald. > > On 02/18/2016 07:01 PM, Ross Walker wrote: >> 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 >> > > -- > ____________________________________________________________________________ > > Prof. Gerald MONARD > SRSMC, Université de Lorraine, CNRS > Boulevard des Aiguillettes B.P. 70239 > F-54506 Vandoeuvre-les-Nancy, FRANCE > > e-mail : Gerald.Monard.univ-lorraine.fr > tel. : +33 (0)383.684.381 > fax : +33 (0)383.684.371 > web : http://www.monard.info > > ____________________________________________________________________________ > > > _______________________________________________ > 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-developersReceived on Thu Feb 18 2016 - 11:30:07 PST