Hi all
Is there the possibility to get the thread identifier inside a parallel region specified with the accelerator directives ? I mean something like the OpenMP runtime function omp_get_thread_num().
In the PGI reference documentation I found nothing.
Thanks
Cheers
A.
Hi A.,
Is there the possibility to get the thread identifier inside a parallel region specified with the accelerator directives ?
Sorry, no.
What would be the use case be for such a feature?
Hi Mat
Thanks for your reply.
What would be the use case be for such a feature?
My goal was to accelerate an OpenMP portion of code through a minimum set of changes using Accelerator directives.
Cheers
A.