I want to pass an array of type variables to cuda kernel. I have a local copy of this array on my host (GLOM) and I want to clone it on my device memory (GLOM_d). For this reason I have declared device type variable as allocatable and after initialized the host variable I try to clone the variables with the following command:
I do not have any error in compilation, but I have a segmentation fault error in running time.
This error is coming out when I have changed some object properties from fixed dimension to allocatable.
Should I change something in my code to clone a type variable with allocatable properties?
Thank you in advance.