This thread is locked.Only browsing is available.
Top Page > Browsing
runtest error
Date: 2016/06/26 12:39
Name: Kazume NISHIDATE   <>

Dear developers,

I'm a beginner of the OpenMX.
I could successfully compile it on my machine.

nisidate@rikyu:~/OpenMX/openmx3.8/work$ uname -a
Linux rikyu 4.4.0-24-generic #43-Ubuntu SMP Wed Jun 8 19:25:16 UTC 2016 i686 i686 i686 GNU/Linux

The compiler and mkl library I use are
which were downloaded from the Intel web-site.
These are supposed to be free for academic use.

The CC and FC lines in the makefile are
CC = icc -O3 -openmp -I/opt/intel/compilers_and_libraries_2016.0.109/linux/mkl/i
nclude/fftw -I/usr/include/mpich
FC = ifort -O3 -openmp -I/usr/include/mpich
LIB= -L/opt/intel/compilers_and_libraries_2016.0.109/linux/mkl/lib -mkl=parallel
-lmkl_intel -lmkl_intel_thread -lmkl_core -lpthread -lifcore -lmpi -lmpif90 -lm

In the testrun, I encountered the error of
nisidate@rikyu: mpirun -np 1 ../source/openmx -runtest

<<< skipping the first output messages >>>

Num. of grids overlapping with atom 1 = 167056
Num. of grids overlapping with atom 2 = 167056

SCF calculation at MD = 1

<MD= 1> Calculation of the overlap matrix
<MD= 1> Calculation of the nonlocal matrix
<MD= 1> Calculation of the VNA projector matrix

******************* MD= 1 SCF= 1 *******************
<Cluster> Solving the eigenvalue problem...
Assertion failed in file src/mpid/ch3/src/ch3u_buffer.c at line 77: FALSE
memcpy argument memory ranges overlap, dst_=0x1431c090 src_=0x1431c090 len_=160

internal ABORT - process 0
[mpiexec@rikyu] handle_pmi_cmd (./pm/pmiserv/pmiserv_cb.c:78): Unrecognized PMI command: abort | cleaning up processes
[mpiexec@rikyu] control_cb (./pm/pmiserv/pmiserv_cb.c:868): unable to process PMI command
[mpiexec@rikyu] HYDT_dmxu_poll_wait_for_event (./tools/demux/demux_poll.c:77): callback returned error status
[mpiexec@rikyu] HYD_pmci_wait_for_completion (./pm/pmiserv/pmiserv_pmci.c:435): error waiting for event
[mpiexec@rikyu] main (./ui/mpich/mpiexec.c:901): process manager error waiting for completion


I also tried the runtime options, "-np2" and "-np4",
but these returned the same message.

Is this a serious error?
Any suggestion is welcome.
Page: [1]

Re: runtest error ( No.1 )
Date: 2016/07/01 21:16
Name: Kazume NISHIDATE  <>

Thank you for your attention.
I'll try some some optional settings, eg. to use Gnu's compilers.

best regards
敬具 西館
Re: runtest error ( No.2 )
Date: 2016/07/04 10:04
Name: T. Ozaki


This message is not due to OpenMX, but something related to your LAPACK, since the place
where the problem happened is calling eigenvalue solvers from LAPACK.



Page: [1]