You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Some code exist in both parallel en sequential form, there are various ways to deal with that, one is the ifdef all the MPI calls, the other is to provide a dummy/empty MPI implementation.
I think it would be a good idea to provide a sequential version of Boost MPI so that users could have both parallel and sequential versions with no source code modifications.
One consequence is that we could provide that version on all platforms, even those with no MPI implementations, as users might still want the sequential version. Which could even help catching some compilations problems.
I expecting some bjam problems though...
The text was updated successfully, but these errors were encountered:
Some code exist in both parallel en sequential form, there are various ways to deal with that, one is the ifdef all the MPI calls, the other is to provide a dummy/empty MPI implementation.
I think it would be a good idea to provide a sequential version of Boost MPI so that users could have both parallel and sequential versions with no source code modifications.
One consequence is that we could provide that version on all platforms, even those with no MPI implementations, as users might still want the sequential version. Which could even help catching some compilations problems.
I expecting some bjam problems though...
The text was updated successfully, but these errors were encountered: