ABINIT Discussion Forums

Dear Abiniters,

I do some gamma point phonon calculation. Abinit runs well with ecut 60 Ha but
stops when increasing to ecut 70 Ha with the following error in the phonon perturbation dataset.

Abinit 6.0.3 has been compiled with ifort and uses openmpi
Same problem does not occur in the sequential version.

Thanks to all

Tanju Gürel
Namık Kemal University
Tekirdag-Turkey

==============================================================================
scfcv3, nstep= 100
scfcv3, nstep= 100
scfcv3, nstep= 100
scfcv3, nstep= 100

ITER STEP NUMBER 1
-P-0000 leave_test : synchronization done...
vtorho3: loop on k-points and spins done in parallel
vtorho3 : MPI_ALLREDUCE, buffer of size10610232 bytes
--------------------------------------------------------------------------
mpirun noticed that process rank 5 with PID 32691 on node canakkale exited on signal 11 (Segmentation fault).
--------------------------------------------------------------------------

Hello again,

I have compiled abinit603 with gfortran and it worked. It is said
to be gfortran is slower than ifort.

Any idea why my problem occurs in ifort?

Thanks
Tanju

Dear all,

Thanks to the post of forum member mancini
http://forum.abinit.org/viewtopic.php?f=17&t=257&start=0

I compiled IFORT with the flags:
FCFLAGS_EXTRA="-heap-arrays 1024"
CFLAGS_EXTRA="-heap-arrays 1024"

and the problem is solved.
Sorry to take your time.

Best,
Tanju