[CPMD-list] CPMD bug???
Matteo Guglielmi
matteo.guglielmi at epfl.ch
Mon Sep 11 20:54:31 CEST 2006
Hello,
I've compiled cpmd-3.11.1 on IBM BlueGene without problems
but when I try to do a Konn-Sham analysis on my system it
crashes all the time in a systematic way.
I noticed that "tuning" the box size can solve the problem.
(this does not mean make it necessarily smaller)
#########################
Here comes my input file:
#########################
&CPMD
RESTART WAVEFUNCTION COORDINATES LATEST
KOHN-SHAM ENERGIES
6
RHOOUT BANDS
8
-38 -39 -40 -41 -42 -43 -44 -45
CONVERGENCE ORBITAL
1.0D-6
LANCZOS PARAMETER
3000 8 39 1.D-8
MIRROR
&END
&DFT
NEWCODE
FUNCTIONAL PBE
&END
&SYSTEM
POISSON SOLVER TUCKERMAN
ANGSTROM
CHARGE
0.0
SYMMETRY
0
CELL ABSOLUTE
25.0 25.0 25.0 0.0 0.0 0.0
CUTOFF
70.0
&END
&ATOMS
*H_MT_PBE KLEINMAN-BYLANDER
LMAX=S
16
12.915 15.410 12.552
13.158 13.089 12.449
9.372 15.187 12.499
8.150 13.001 12.503
8.265 10.464 12.578
9.706 8.423 12.559
12.189 8.641 12.490
14.808 14.711 13.798
15.091 16.071 12.941
14.991 14.617 12.131
16.638 13.162 10.776
15.165 12.858 10.155
15.468 12.484 11.750
14.961 11.534 14.547
15.662 10.581 13.414
14.103 11.092 13.179
*N_MT_PBE KLEINMAN-BYLANDER
LMAX=P
4
12.201 10.700 12.478
14.579 15.181 12.918
15.638 13.183 11.002
15.032 11.384 13.534
*C_MT_PBE KLEINMAN-BYLANDER
LMAX=P
9
11.341 14.263 12.450
12.066 13.072 12.473
11.412 11.816 12.473
9.969 11.816 12.473
9.246 13.032 12.481
9.910 14.225 12.483
11.561 9.541 12.490
10.152 9.417 12.525
9.356 10.546 12.532
*O_MT_PBE KLEINMAN-BYLANDER
LMAX=P
1
11.893 15.488 12.372
&END
#########################################
Here comes the **console** error message:
#########################################
Rzv:cannot allocate unexpected buffer from R:7 T:1 C:0
Dumping 16 frames
Frame 0: 0x5c003c
Frame 1: 0x5c2acc
Frame 2: 0x5e6cd0
Frame 3: 0x5e075c
Frame 4: 0x5e2c28
Frame 5: 0x5c0258
Frame 6: 0x5ba41c
Frame 7: 0x5a3a14
Frame 8: 0x4a1efc
Frame 9: 0x2b33c0
Frame 10: 0x27d890
Frame 11: 0x331b24
Frame 12: 0x330fb8
Frame 13: 0x2008c8
Frame 14: 0x200520
Frame 15: 0x20016c
Posted Queue:
-------------
Posted Requests 0, Total Mem: 0 bytes
Unexpected Queue:
-----------------
Unexpected Requests 235, Total Mem: 189799444 bytes
Fatal: Cannot allocate buffer for unexpected message<Sep 11 20:45:14.236438>
BE_MPI (ERROR): The error message in the job record is as follows:
<Sep 11 20:45:14.236524> BE_MPI (ERROR): "killed by exit(1) on node 0"
<Sep 11 20:45:14.494342> BE_MPI (ERROR): The error message in the job record is
as follows:
<Sep 11 20:45:14.494373> BE_MPI (ERROR): "killed by exit(1) on node 0"
###################################################
Here comes the log file (**no error messages**!!!):
###################################################
.
.
.
(R) ESR = 6.04421677 A.U.
(X) EXCHANGE-CORRELATION ENERGY = -34.98520382 A.U.
(V) EXCHANGE-CORRELATION POTEN. = -45.43107648 A.U.
GRADIENT CORRECTION ENERGY = -.58398329 A.U.
==------------------------------------------------------------==
== NFI= 1 ETOT= -116.451443 TCPU= 17.58 ==
== DRHOMAX= 4.328E-04 DETOT= 0.000E+00 THL= 0.000E+00 ==
==------------------------------------------------------------==
PS: The geometry file is in the attachment.
Thanks,
MG.
+---------------------------------------------------------+
| Address ----------------------------> Matteo Guglielmi |
| EPFL-SB-ISIC-LCBC |
| Ecole Polytechnique BCH 4201 |
| Federale de Lausanne CH-1015 Lausanne |
| |
| Phone --> +41-(0)21-693-0327 _____ |
| Fax --> +41-(0)21-693-0320 | | |
| Mobile1 --> +41-(0)76-5245216 ------- |
| Mobile2 --> +39-340-2751607 0-0 |
| Web --> http://lcbcpc21.epfl.ch c |
| EMail1 --> matteo.guglielmi at epfl.ch +--o00o-----o00o--+
| EMail2 --> matteo.guglielmi at pnl.gov |ICanSeeClearlyNow|
| ICQ --> #166904425 |...TheBrainIsGone|
+---------------------------------------+-----------------+
-------------- next part --------------
A non-text attachment was scrubbed...
Name: GEOMETRY
Type: application/octet-stream
Size: 3870 bytes
Desc: not available
Url : http://cpmd.org/pipermail/cpmd-list/attachments/20060911/3f1b4ec6/attachment.obj
More information about the CPMD-list
mailing list