ABINIT Discussion Forums

Hello every one,
When I want to calculate the linear and non-linear optical properties by:
ngkpt 24 24 24
nband1 10
nband2,3,4,5,6 50
ecut 50
I reach the following error:
Error: overflow of mcg integer for size of the full wf. Recompile with large int or reduce system size.
Thanks a lot for your help and time.
Asadi Yaser

You have to put CFLAGS , CXXFLAGS, FCFLAGS to the size of integer you want.

If you use gfortran
-fdefault-integer-8
if you use ifort
-i8

The option of your compiler can be know with :
man gfortran or gfortran --help
man ifort or ifort --help

To know if under your mpif90, mpicc ect is ifort, gfortran or any other compiler do

mpif90 --help for instance

Hope it will help

Hello,
It means that the maximum integer defined is not sufficient for your case.
You can set option to the compiler to increase the maximum integer.
You have first to know which compiler are under mpif90 (gfortran , ifort ...), mpicc (gcc, icc..).
For that mpif90 --help , mpicc --help might give you an answer.
Each compiler gfortran, ifort , ... has an option to stare the size of integer ( which stare the maximum integer).
For instance :
ifort -i8 : set default KIND of integer and logical variables to 8 ( defaut is 4 not enought)
gfortran -fdefault-integer-8 : Set the default integer and logical types to an 8 byte wide type ( default is 4)
The option of compilers can be seen by man ifort, man gfortran or ifort--help, gfortran --help
To stare these options to compile abinit
use in the configure the options CFLAGS, FCFLAGS ou CXXFLAGS

Hi Dear delaveau, thank you very much.
Best regards.