[Pw_forum] compiled pw, cp on a BG/P - scf.in test hangs

J. J. Ramsey jjr19 at uakron.edu
Wed Dec 24 19:30:15 CET 2008


______________________________

So I was able to get pw and cp to compile on the BG/P, but when I try and test out pw using the scf.in test, the code hangs at the following (regardless of number of procs, time, etc)

launched as follows: 

Program PWSCF     v.4.0.4  starts ...
     Today is 24Dec2008 at 12: 7:44 

     Parallel version (MPI)

     Number of processors in use:       1

     For Norm-Conserving or Ultrasoft (Vanderbilt) Pseudopotentials or PAW

     Current dimensions of program pwscf are:
     Max number of different atomic species (ntypx) = 10
     Max number of k-points (npk) =  40000
     Max angular momentum in pseudopotentials (lmaxx) =  3

(hangs here until killed)

Any ideas what could cause this? Maybe a missed option someplace? I don't get any useful information sent to stderr

-----------------------------

My guess is that something goofy happened with input redirection. If you haven't done it already, try running the test as "mpirun-or-exec ./pw.x -in scf.in > scf.out", and see what you get.

(BTW, I've noticed that on a cluster that I've used, running the "check-pw.x.j" within a batch job works fine even though input redirection in the script uses the "<" operator, but when running mpirun.lsf directly from the batch script, I need the "-in" option. Go figure.)



      



More information about the users mailing list