Firefly and PC GAMESS-related discussion club


 
Learn how to ask questions correctly  
 
 
We are NATO-free zone
 



Re: Dump of registers follows (caught signal 11)

Alex Granovsky
gran@classic.chem.msu.su


On Tue Mar 14 '17 5:56pm, Andrey Degtyarev wrote
------------------------------------------------
>Dear colleagues,
>When i run FF, get a memory error caught signal 11.
>Firefly version 8.1.1, build number 9295
>OS Ubuntu 16.04 kernel 4.8.0-36-generic
>The problem is also reproduced in Debian 8.0 and different architecture the CPU.
>Input and dupm files in attach file

Dear Andrey,

Running your input with the latest Firefly version 8.2.0, Firefly
gracefully aborts with the following error message:

THE ECP RUN REMOVES  150 CORE ELECTRONS, AND THE SAME NUMBER OF PROTONS.
THE ADJUSTED NUCLEAR REPULSION ENERGY=      962.9513732419

ERROR: NEGATIVE OVERALL NUMBER OF ELECTRONS!

ADDRESS 0x0098A2F9 HAS INITIATED PROGRAM ABORT BECAUSE OF FATAL ERROR(S)

The check for negative number of electrons was absent in the earlier
versions of Firefly and I suppose the negative number of electrons
is the reason of crash (page fault) you have observed with Firefly
version 8.1.1

The underlying problem is that present ECP code removes core electrons
even if the corresponding nuclear charge is negative, i.e. from the ghost
atoms. You need to modify Ti-ecp in your input file as follows.

Replace

Ti-ecp GEN   10     2

by

Ti-ecp GEN    0     2

If there were a mixture of ghost and non-ghost Ti atoms in your
input structure, you would need to define two ECPs for Ti atoms
of course using different names for them; and use these ECPs
appropriately. The first ECP would remove zero core electrons
and the second ECP would remove 10 electrons, being otherwise
identical to the first one.

Kind regards,
Alex Granovsky






[ Previous ] [ Next ] [ Index ]           Sat Mar 18 '17 3:25pm
[ Reply ] [ Edit ] [ Delete ]           This message read 136 times