[Wien] lapw1(c)_mpi crash in the initialization procedure

Florent Boucher Florent.Boucher at cnrs-imn.fr
Wed Mar 16 10:44:37 CET 2005


Dear Wien users,
I succed in compiling the lapw1_mpi package using the blacs (compaq 
supported) and cxml library. I am running the latest WIEN2K_05.2 on 
alphaEV7 computer (TRU64).
I have a crash when I try to start with the benchmark (also the same is 
true with other exmaples).
Below is the output of the crash. I was able to run the WIEN2K_03 
version using MPI but since the new modification in the modules.F part 
concerning the HNS, the program crash here

! for HNS: row-wise blockcyclic distribution over npe processors
->>      call blacs_gridinfo(ICTXTALLR,npe, 1, myrow, mycol)
      lld = max(1,numroc(N, blocksize, myrow, 0, npe))
      call descinit(descr, N, NSLMAX*NSLMAX,         &
                    blocksize, NSLMAX*NSLMAX, 0, 0, ICTXTALLR, lld, INFO)
      print *, myrow, '> nsl2 = ', nslmax*nslmax, ', lld = ', lld
      print *, 'descr = ', descr

Does anybody (may be M.Dohmen) has an idea or is already able to run 
lapw1_mpi on such a computer.

I have tried the simple Hello program from blacs (that use the same 
subroutines) and it works perfectly.
      PROGRAM HELLO
*     -- BLACS example code --
*     Written by Clint Whaley 7/26/94
*     Performs a simple check-in type hello world
*
Thank you for you help.
Florent



boucher at phoenix - 163% dmpirun -np 2 $WIENROOT/SRC_lapw1/lapw1c_mpi 
lapw1.def
 Using            2  processors, My ID =            1
 Using            2  processors, My ID =            0
forrtl: severe (174): SIGSEGV, segmentation fault occurred
forrtl: severe (174): SIGSEGV, segmentation fault occurred
   0: __FINI_00_remove_gp_range [0x3ff81a6de38]
   1: __FINI_00_remove_gp_range [0x3ff81a77414]
   2: __FINI_00_remove_gp_range [0x3ff800d9c30]
   3: blacs_gridinfo_ [0x1201575f4]
   0: __FINI_00_remove_gp_range    4: $parallel$init_parallelmatrices_ 
[0x3ff81a6de38]
   1: __FINI_00_remove_gp_range [0x3ff81a77414]
   2: __FINI_00_remove_gp_range [0x3ff800d9c30]
   3: blacs_gridinfo_ [0x1201575f4]
   4: $parallel$init_parallelmatrices_ [modules_tmp_.F: 121, 0x12004554c]
   5: $matrices$init_matrices_ [modules_tmp_.F: 121, 0x12004554c]
   5: $matrices$init_matrices_ [modules_tmp_.F: 495, 0x12004a2bc]
   6: coors_ [modules_tmp_.F: 495, 0x12004a2bc]
   6: coors_ [coors.f: 201, 0x12001d03c]
   7: calkpt_ [coors.f: 201, 0x12001d03c]
   7: calkpt_ [calkpt_tmp_.F: 142, 0x1200196cc]
   8: lapw1_ [calkpt_tmp_.F: 142, 0x1200196cc]
   8: lapw1_ [lapw1_tmp_.F: 51, 0x1200416ec]
   9: main [lapw1_tmp_.F: 51, 0x1200416ec]
   9: main [for_main.c: 203, 0x12007547c]
  10: __start [0x120011a48]
[for_main.c: 203, 0x12007547c]
  10: __start [0x120011a48]
MPI process 591024 exited with status 174
MPI process 593346 exited with status 174

-- 
 -------------------------------------------------------------------------
| Florent BOUCHER                    |                                    |
| Institut des Matériaux Jean Rouxel | Mailto:Florent.Boucher at cnrs-imn.fr |
| 2, rue de la Houssinière           | Phone: (33) 2 40 37 39 24          |
| BP 32229                           | Fax:   (33) 2 40 37 39 95          |
| 44322 NANTES CEDEX 3 (FRANCE)      | http://www.cnrs-imn.fr             |
 -------------------------------------------------------------------------




More information about the Wien mailing list