Custom Search
-- 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-developersReceived on Wed Feb 17 2016 - 17:30:03 PST