On Sat, 24 Apr 2010, Boyce Griffith wrote:
> It appears that one problem is that the innermost parallel_reduce is
> being called within a parallel_for. Regardless of the number of
> threads, calls to parallel_XXX call BoolAcquire(in_threads), and so the
> innermost call to parallel_reduce result
It appears that one problem is that the innermost parallel_reduce is
being called within a parallel_for. Regardless of the number of
threads, calls to parallel_XXX call BoolAcquire(in_threads), and so the
innermost call to parallel_reduce results in an assertion failure.
I'm not sure that this
Hmmm this might have to do with some changes I made. I was having
trouble instantiating the PointLocator class in parallel for use with
periodic BCs with threading and put some guards around it's
creation... although it shouldn't be a problem if you're not actually
using threads...
I'
Hi, Folks --
I am getting an assertion failure when calling EquationSystems::init()
which which seems to be related to threading support when I add a
periodic boundary to a simple rectangular mesh. Everything seems to run
fine when I don't add the periodic boundary. I am not using multiple
t
