[molpro-user] FCI problem (molpro 2012.1)

[Jeff Hammond]

That Armci error is a segfault related to comm buffers. I can't say
any more than that though. You might try the MPI build ie no-ARMCI
build.

Jeff

Sent from my iPhone

On Oct 31, 2013, at 3:24 AM, Jussi Eloranta <jmeloranta at gmail.com> wrote:

> Hi,
>
> I am trying to run a fairly large full CI calculation and keep getting segmentation faults (hostname removed; XXX):
>
> 0:Segmentation Violation error, status=: 11
> (rank:0 hostname:XXX pid:61421):ARMCI DASSERT fail. src/common/signaltrap.c:SigSegvHandler():310 cond:0
> 3:Segmentation Violation error, status=: 11
> (rank:3 hostname:XXX pid:61424):ARMCI DASSERT fail. src/common/signaltrap.c:SigSegvHandler():310 cond:0
> Last System Error Message from Task 0:: Inappropriate ioctl for device
> application called MPI_Abort(comm=0x84000007, 11) - process 0
> Last System Error Message from Task 3:: Inappropriate ioctl for device
> application called MPI_Abort(comm=0x84000004, 11) - process 3
> 1:Segmentation Violation error, status=: 11
> (rank:1 hostname:XXX pid:61422):ARMCI DASSERT fail. src/common/signaltrap.c:SigSegvHandler():310 cond:0
> Last System Error Message from Task 1:: Inappropriate ioctl for device
> application called MPI_Abort(comm=0x84000004, 11) - process 1
>
> ===================================================================================
> =   BAD TERMINATION OF ONE OF YOUR APPLICATION PROCESSES
> =   EXIT CODE: 11
> =   CLEANING UP REMAINING PROCESSES
> =   YOU CAN IGNORE THE BELOW CLEANUP MESSAGES
> ===================================================================================
>
>
> ---
>
> My input file is as follows:
>
> ***,He_2^{*-}
> memory,4000,m
>
> r = [1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 7.0 8.0 9.0 10.0 12.0 14.0 16.0 18.0 20.0 25.0 30.0 40.0 50.0 100.0 200.0]
>
> sexp=[     720.415859,     108.159574,      24.672815, 7.008060,       2.283030,       0.809432,       0.296071, 0.088547,       0.033435,       0.009982]
> pexp=[       6.681348,       1.975533,       0.670324, 0.174403,       0.049119,       0.014593,       0.003627]
> dexp=[       7.336700,       2.521632,       0.903599, 0.194153,       0.011633,       0.043237]
>
> basis={
> s,He,sexp(1),sexp(2),sexp(3),sexp(4),sexp(5),sexp(6),sexp(7),sexp(8),sexp(9),sexp(10)
> c,1.1,1.0
> c,2.2,1.0
> c,3.3,1.0
> c,4.4,1.0
> c,5.5,1.0
> c,6.6,1.0
> c,7.7,1.0
> c,8.8,1.0
> c,9.9,1.0
> c,10.10,1.0
> p,He,pexp(1),pexp(2),pexp(3),pexp(4),pexp(5),pexp(6),pexp(7)
> c,1.1,1.0
> c,2.2,1.0
> c,3.3,1.0
> c,4.4,1.0
> c,5.5,1.0
> c,6.6,1.0
> c,7.7,1.0
> d,He,dexp(1),dexp(2),dexp(3),dexp(4),dexp(5),dexp(6)
> c,1.1,1.0
> c,2.2,1.0
> c,3.3,1.0
> c,4.4,1.0
> c,5.5,1.0
> c,6.6,1.0
> }
>
> geometry={
> He1
> He2, He1, r(i)
> }
>
> do i = 1,#r
>
> {hf; wf,5,5,3,-1; closed,1;open,1.5,2.1,3.1}
> {fci;core,;occ,24,12,12,5,24,12,12,5;wf,5,5,3,-1;state,1}
> e1(i)=energy(1)
>
> enddo
>
> table,r,e1
>
> ---
>
> In the above example, I redcued the active space but it still core dumps. If I reduce it further, the calculation runs OK.
> Thus, I suspect that this has something to do with memory allocation. Here are couple of observations that may help:
>
> 1) If I request too much memory, I will see the out of memory termination message in the system log (dmesg). However, with the above input, I don't see that.
> So the system is not running out of memory.
>
> 2) If I request only about 800mw or so, molpro will complain that not enough memory allocated. So, molpro is apparently happy with the amount of memory requested.
>
> 3) I get the segmentation fault even if run with just one processor.
>
> Any clues what might be happening?
>
> Thanks,
>
> Jussi Eloranta
>
>
>
> _______________________________________________
> Molpro-user mailing list
> Molpro-user at molpro.net
> http://www.molpro.net/mailman/listinfo/molpro-user