Custom Search
-- 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 - 13:30:03 PST