Re: Allocatables in derived types and functions

Thanks! I am looking forward to the new compiler.

Note that the code runs if the contents of dummy is changed from 2 to 1
allocatable array! It still has to check to see if the allocatable
component need to be deallocated. So, this test should also go wrong
then, right?! How can it then be that the code runs in that case?.
Might it be that there is another issue as well?

BTW. If I encounter something peculiar with the compiler, is there any
way to communicate that directly to intel, rather than through this
news group? Mind that I am a mere user and have no rights with respect
to installing etc.; that is the job of our system administrators.

Thanks again

Arno

Steve Lionel wrote:
Arno wrote:

The latest version of the compiler will be installed, but may take some
time. So, I will post the code, just to be sure.

Thanks for the code. The current compiler still has a problem with it.
What is happening is that when write_i is called, because dummy is
intent(out), it has to check to see if the allocatable components of
dummy need to be deallocated and do so if they are. They aren't, but
it is getting this test wrong and trying to deallocate memory that
isn't allocated. We'll fix this.


Steve Lionel
Developer Products Division
Intel Corporation
Nashua, NH

User communities for Intel Software Development Products
http://softwareforums.intel.com/
Intel Fortran Support
http://developer.intel.com/software/products/support/
My Fortran blog
http://www.intel.com/software/drfortran

.