AMBER Archive (2003)

Subject: Linux cluster Amber7 sander mpi error -- Null communicator, IOT Trap

From: Chris Switzer (switzer_at_mail.ucr.edu)
Date: Fri Jan 03 2003 - 21:18:11 CST

Amber 7 sander newly compiled on a linux cluster gives the following error:

1 - MPI_COMM_RANK : Null communicator
[1] Aborting program !
[1] Aborting program!
Process aborting...
IOT Trap
0 - MPI_COMM_RANK : Null communicator
[0] Aborting program !
[0] Aborting program!
Process aborting...
IOT Trap
-------------------

System:
      Linux cluster 2.4.9-31smp
       Mpich-1.2.1..7b
-------------------

Makefile used to create the Amber7 sander giving above error:
      An altered version of "Machine.g77_mpich". Amber7 would only finish compiling when the "g77" references in Machine.g77_mpich were changed to "mpif77" per a reflector e-mail.
-------------------

It is noteworthy that when Amber7 is compiled non-parallel using Machine.g77 without any alterations, sander runs fine.
-------------------
-------------------

Some additional notes...
Compiling behavior with other machine files:
       
Attempted compiling with Machine.g77_mpich unaltered gives the following type of error:
         g77 -c -g _nxtsec_.f
        ../Compile LOAD -o new2oldparm new2oldparm.o nxtsec.o
        g77 -O6 -o new2oldparm new2oldparm.o nxtsec.o -lm -L/usr/local/mpich-1.2.4..8/lib -lmpich
        /usr/local/mpich-1.2.4..8/lib/libmpich.a(gmpi_regcache.o): In function `gmpi_regcache_init':
       gmpi_regcache.o(.text+0x1e): undefined reference to `gm_hash_hash_ptr'
       gmpi_regcache.o(.text+0x23): undefined reference to `gm_hash_compare_ptrs'
      etc....
        make[1]: *** [new2oldparm] Error 1
       make[1]: Leaving directory `/home/switzer/amber7/src/lib'
       make: *** [install] Error 2

 I have pgf77. Attempted compilation with Machine.pgf77_mpi gives errors of the following sort:
       SYSLIB=`../sysdir lib` ; ../Compile LOAD -o sander \
        sander.o ....etc..... ../blas/blas.a ../lib/nxtsec.o $SYSLIB;
         pgf77 -o sander sander.o cshf.o ....etc..... decomp.o ../lapack/lapack.a ../blas/blas.a ../lib/nxtsec.o
          /home/switzer/amber7/src/Machines/standard/sys.a -lm
           sander.o: In function `trajene':
        _sander_.f:1084: undefined reference to `mpi_init__'
        _sander_.f:1085: undefined reference to `mpi_comm_rank__'
        _sander_.f:1086: undefined reference to `mpi_comm_size__'
          _sander_.f:1282: undefined reference to `mpi_bcast__'
          ....etc....
         new_time.o(.text+0x2f46): undefined reference to `mpi_send__'
           new_time.o(.text+0x32dd): undefined reference to `mpi_recv__'
             make[1]: *** [sander] Error 1
             make[1]: Leaving directory `/home/switzer/amber7/src/sander'
            make: *** [install] Error 2
-------------------

Any help much appreciated.

Sincerely,

Chris Switzer
Chemistry Dept
UC Riverside
Riverside, CA
92521